Strontium-90 Adjacent to Fall Chinook Salmon Redds 

NORM BUSKE / Government Accountability Project 25jun99

Strontium-90 Adjacent to Fall Chinook Salmon Redds at H-Reactor Area of the Hanford Reach of the Columbia River

Strontium 90 intrusion, at ten times permissible activity, from the Hanford H Reactor area into the Columbia Riverbed at a very local area of Fall Chinook spawning grounds is here reported for the first time. This discovery raises concerns for the health of the salmon stock and for the adequacy of governmental regulatory oversight of large scale, dangerous activities of the U.S. Department of Energy at Hanford Site.

Introduction 

Fall chinook salmon (Oncorhynchus tshawytscha) spawn in the rocky bed of the Columbia River next to the decommissioned H-Reactor Area of the Department of Energy Hanford Site. See Fig. 1 for general location. These salmon are an important regional resource for commercial, tribal, and sport fisheries [1] 

According to Glen Spain, Conservation Director of the Institute for Fishery Resources, up to 80% of the fall chinook salmon that spawn in the Columbia River spawn along the Hanford Reach of the river. Early life stages --egg, alevin, and fry-- of these salmon are particularly susceptible to local contaminant exposure, because these stages remain within the bottom substrate at the nest for about three months [2]. 

Hexavalent chromium released from historic nuclear weapons production processes on the site have entered groundwater which upwells from some areas where fall chinook salmon dig nests in the riverbed and spawn [3]. 

These salmon redds have been observed within 50-100 feet of the H-Reactor Area shoreline where hexavalent chromium in riverbed water has been measured at 52-130 microgram per liter (µg/L). Hexavalent chromium above 11µg/L has been identified as a "contaminant of concern" for the non-motile, early life stages of the fall chinook salmon adjacent to the H-Reactor Area [4]. 

Consequently, the U.S. Department of Energy has undertaken interim remedial measures at H-Reactor Area. Studies contracted by the Department of Energy have provided no evidence of any adverse effects to fall chinook salmon from radioactive materials. This assurance is based on non-detection of any artificial radionuclides in river water or riverbed water even approaching drinking water standards established by the Washington State and the Environmental Protection Agency [5]. 

The salmon redds extend from within about 100 feet of the H-Reactor Area shoreline out into deeper water, Fig. 2 [6]. 

The greatest concentration of hexavalent chromium (up to 130µg/L) in riverbed water in both shallow and deep water were found by Hope and Peterson on their Transect 1, at the top of Fig. 3 [7]. 

These high concentrations of toxic contaminant are presumably associated with a riverbed seep comparable to shoreline seeps 152-2 and 153-1 marked in Fig. 2. Previous work by the present author has characterized such localized areas of seepage --springs-- as the outpourings of narrow, high-speed, low-sorption groundwater pathways. Hope and Peterson's measurements of hexavalent chromium in riverbed water at concentrations equal to (or even above) the highest monitoring well values, along a single transect is consistent with this characterization. Groundwater measurements reported from annual monitoring of Hanford Site have been depicted as several "plumes" of strontium-90 (Sr-90) at old reactor areas next to the Columbia River shoreline [8]. 

One of these Sr-90 plume portrayals is in the H-Reactor Area next to the shoreline, Fig. 4 [9]. 

Based on the co-existence of elevated Sr-90 in groundwater with a localized, contaminant seepage offshore, the author hypothesized that groundwater contaminated with both hexavalent chromium and Sr-90 is following a single narrow "underground stream" pathway from a source in H-Reactor Area to Transect 1 in salmon redds offshore. This hypothesis has been tested by seeking to locate a narrow Sr-90 path crossing the H-Reactor Area near the shore-end of Transect 1. This Sr-90 pathway has been located by detection of Sr-90 in leaves of mulberry trees on the H-Reactor Area shoreline by means of a survey geiger counter with a deck of 6 leaves placed against the detector. Strontium-90 was then analyzed in samples of dried mulberry leaves, by means of shape fitting the energy spectrum counted on a sodium-iodide, well detector operating in the 30-1500 kiloelectron volt (KeV) photon energy range to quantify results. Conceptual Model The author has previously shown that contaminant plume models based on monitoring well samples are unrealistic [10]. 

They do not portray high-speed, low-sorption pathways which are line-like features that end in narrow shoreline and riverbed seeps that are seen along the Hanford shore when the Columbia River is at a low stage and are underwater at higher river levels [11]. 

The question of portrayal of Hanford groundwater migration as "plumes" or as "underground streams" has been partly reviewed by the U.S. Geological Survey [12]. 

Although some seepages of groundwater might be better characterized as "plumes" than as "underground streams," the underground stream characterization describes situations in which contaminant traveltimes are unusually low and concentrations surprisingly high. In the present case, the highest hexavalent chromium concentration in the H-Area salmon redds --130µg/L-- is actually higher than the highest concentration reported in any well onshore --110µg/L-- Fig. 3. This curiosity is readily explained by the "underground stream" characterization: None of the wells happen to tap into the underground stream at H-Reactor Area. A narrow seepage (Site #3) of Sr-90 laden groundwater at N-Reactor Area has been studied for years [13]. 

The author picked mulberries from a tree at the N Springs shoreline in 1990 and prepared jam to call attention to the biological pathways of Sr-90 in this seepage. Beginning in September 1998, the author has collected mulberry leaves from trees in this seepage, examining the relationship between the Sr-90 readily and consistently sampled and detected in these leaves, river stage and season, and Sr-90 reported in water in wells monitoring Site #3 at N-Springs. 

This sampling has been thwarted by Department of Energy efforts , supported by the Washington State Department of Ecology, to clear N-Springs of radioactive mulberry trees in a public access area. On June 11, 1999 contractors were observed chipping cleared mulberry trees from the N-Reactor Area. That sample of mulberry leaves collected from Site #3 yielded 202,000 picocuries per dry kilogram [pCi/kg (dry)] of Sr-90 in comparison to a nearby groundwater monitoring well average of 5,900 pCi/L [14]. 

At that time and high river stage, the annual average groundwater Sr-90 activity (in pCi/L) was estimated to be 0.029 times the sampled activity in dried mulberry leaves [in pCi/kg(dry)]. (Inasmuch as the mass of a liter of water is one kilogram, this factor of 0.029 is practically dimensionless.) This factor of 0.029 = average groundwater / dried mulberry leaves on June 11 for Sr-90 between groundwater and mulberry leaves at N-Springs was assumed to apply equally between average groundwater in an "underground stream" at H-Reactor Area and mulberry leaves collected from a tree identified as on top of a hexavalent-chromium contaminated "underground stream" leading to salmon redds offshore. 

This conceptual model assumes the hexavalent chromium and the Sr-90 at the H-Reactor Area share an "underground stream" pathway that is located at the Sr-90 maximum in mulberry leaves from trees surveyed on the shoreline. The relevant Sr-90 value is assumed to be the Sr-90 in this maximum-leaves sample. Then consistency and reality checks are applied to confirm the model of Sr-90 co-migration with a hexavalent chromium spike to salmon redds off the H-Reactor Area is sensible. These checks are part of the conceptual model that is discussed in a later section of this report. Results The approximate position where the hexavalent chromium pathway would cross the shoreline bases on groundwater flow in an ESE direction was estimated from the map of Fig. 3. The author picked one sample of mulberry leaves from one candidate tree near the estimated pathway location on the shoreline on each of three days (May 20, May 27, and June 11, 1999). 

These three samples were analyzed for Sr-90. The first of these three leaf samples did not conform to specific gravity of 0.25, which was specified after this sample was counted. This sample count was thus used only qualitatively. The second sample tested negative for Sr-90. The third mulberry leaf sample, collected on June 11 tested positive for Sr-90. The reported result Sr-90 = 2400 pCi/kg(dry) is a differential measurement, with the photon spectrum from the second sample subtracted. This is a conservative analytical procedure which increases precision and confidence in the positive result while biasing the result to under-estimate Sr-90. Decks of mulberry tree leaves (or canary reed grass where no mulberry trees were available) were collected from the H-Reactor Area shoreline between the outfall structure (at GPS nominal 46° 42' 15" North, 119° 28' 41" West) to the downstream end of the line of trees (at GPS nominal 46° 42' 06" North, 119° 28' 37" West) on June 17, 1999. 

These decks of leaves were counted with a Ludlum 44-9 detector. The highest value, 50% above background, was found in leaves from the mulberry tree already sampled on June 11. This survey result indicated the underground stream of high-strontium groundwater is no more than one mulberry tree wide, say 50 feet. This scale of width is consistent with both shoreline seeps observed in the area and elevated hexavalent chromium values in Transect 1 but not in Transect 2, Fig. 3. These results indicate a high-strontium, underground stream passes beneath the upstream mulberry tree in the line of trees beginning 300 feet south of the H-Reactor outfall structure, just north of the "1" indicating Transect #1 in Fig. 3. 

Multiplying this conservatively measured Sr-90 value in dried mulberry leaves by the factor of 0.029 obtained in the conceptual model of the last section, an average groundwater activity for Sr-90 in the underground stream near the H-Reactor Area shoreline is obtained: Calculated average groundwater at shoreline crossing: Sr-90 = 70 pCi/L Referring to the conceptual model of this study, this calculated average Sr-90 in groundwater estimates Sr-90 measurable in replication of Transect #1 hexavalent chromium measurements in Hope and Peterson's study [3]. 

Therefore, Sr-90 is here calculated in riverbed water at fall chinook redds at H-Reactor Area at ~70 picocuries per liter (pCi/L), which is 8 times the drinking water standard of 8pCi/L. Concerns Assuming Sr-90 is co-migrating with hexavalent chromium in an underground stream discharging to the riverbed at Transect #1 in Fig. 3, a preliminary evaluation of level of concern for early life stages of locally spawned, fall chinook salmon is based on comparison to maximum hexavalent chromium measured at the redds. For hexavalent chromium, the applicable standard is the Environmental Protection Agency's ambient water quality criterion; for Sr-90, the corresponding criterion is the interim drinking water standard. 

These are compared preliminarily in the upper half of the following table, assuming a shared underground stream, based on the considerations already presented: hexavalent chromium strontium-90 maximum in Transect #1: 130µg/L (measured) 70pCi/L (calculated) relevant standard: 11µg/L 8pCi/L over- standard: 12X (measured) 8X (calculated) maximum nearshore well: 110µg/L (measured) 45pCi/L (measured) relevant standard: 11µg/L 8pCi/L over-standard: 10X (measured) 6X (measured) An initial investigation to identify contaminants of concern would include both nearshore well concentrations above relevant standards and concentrations measured at environmental points of ecological concern. The above tabulation shows the considerations of nearshore well data and riverbed measurements and/or calculations yield comparable results. In this quantitative sense, maximum hexavalent chromium intensity is quantified to be about half again as great as Sr-90 intensity: 10-12X standard versus 6-8X standard. Next, an initial investigation to valuate concern would consider actual exposure of potentially susceptible life stages to contaminants of possible concern. In this regard, the fall chinook redds pattern at D-Reactor Area can be compared to the redds pattern at H-Reactor Area. 

The relevant consideration is that the releases of hexavalent chromium to D-Reactor Area and to the upstream portion of the H-Reactor Area redds is not accompanied by Sr-90; whereas, there is Sr-90 in Transect #1 at H-Reactor Area. This comparison reveals the only reported location of salmon redds in substrate hexavalent chromium above 11µg/L is in Transect #1 at H-Reactor Area. This introduces a concern either that spawning salmon might be attracted to Sr-90 contaminated areas, or their might be a compounding effect. This simple consideration of impact location multiplies preliminary concern for potential consequences of Sr-90 contamination at fall chinook salmon redds in the Hanford Reach of the Columbia River. Strontium-90 is therefore anticipated to be of greater ecological consequence to fall chinook salmon than hexavalent chromium in the Hanford Reach spawning grounds. Determination of absolute levels of concern is beyond the scope of this report. However, the concern associated with hexavalent chromium releases to the Columbia River has been sufficient to implement costly remediation efforts. 

Therefore, a semi-quantitative risk assessment of Sr-90 on the Hanford Reach population of fall chinook salmon is a high priority to assure the future quality of an important resource. Validation Strontium-90 plumes exceeding the 8pCi/L drinking water standard are depicted in the annual site monitoring report, up to the shoreline at the B- , K-, N-, H-, and F-Reactor Areas along the Hanford Reach [15]. 

The result of the present study, that at least one of these above-standard strontium contaminations intrudes into salmon redds is of immediate public concern for an important resource in the Northwest. In one practical sense, a potential concern for Sr-90 in groundwater seeps emerging into redds the riverbed at H-Reactor Area is obvious from Fig. 4. The 8pCi/L contour line has higher values on the river side than on the land side of the line, immediately suggesting the potential for high Sr-90 activities where groundwater enters the river. This contouring of the Sr-90 data at H-Reactor Area, together with the above-standard activities measured in nearshore wells implies the presence of Sr-90 emerging from the riverbed. The technical question is: 

What are representative activities of Sr-90 at salmon redds. To answer this question by means of a small field study, an appropriate indicator for sampling --mulberry trees yielding leaves which uptake Sr-90-- and a quick laboratory analysis to allow rapid study iteration are essential. Benchmark analysis for this study is provided by a well-type sodium-iodide photon detector operating in the 20-1500 KeV energy region. This detector is emplaced in a copper-lined, pure lead shield. Important, detrimental nonlinearities of this system are minimized by thermostasis to 0.05°C; by single-point spectral stabilization at 661.7 KeV by means of a cesium-137 source; by imposition of limits on sample activities, activity-relaxation times, and densities; and by extensive use of subtraction techniques. Standard specific gravity of samples for this study is 0.25, with cosmetic puffs used to lighten dense solid media and sorb liquid standards. This density is easily achieved by crushing dried mulberry leaves in an expendable plastic bag, more or less finely, and then packing the crushed leaves more or less firmly into the 125mL sample bottles for which the well detector was designed. Routine sample counting times are 3000 minutes (two days), which limits the sample handling capability of the study. 

At this count time, errors are primarily systematic, and these are biased to under-estimate Sr-90 at low levels, and then the biases are minimized to assure confidence in detection at the stated activity levels to satisfy the special purposes of the present study. Sample spectra are acquired into 4000 channels, according to energy. These raw energy spectra are then transformed into 250 channels for constant-photo-peak width spectra. The photo-peaks are about 8 channels wide (at inflection point), regardless of energy. This feature enhances expert control over the analytical process. This instrument yields a three-lobed, characteristic, broad-band photon spectrum for Sr-90 in dried vegetation samples. No chemistry is required. For Sr-90 activities >10pCi/g(dry), the analytical technique involves subtraction of a blank (12,000 minute count). 

Then a potassium-40 spectrum is subtracted, based on the count of the 1460.8 KeV gamma peak. For Sr-90 <10 pCi/g(dry), a very local background sample is subtracted from a matched candidate sample, for example, leaves from two different mulberry trees close together. This close matching subtracts out most interfering effects. Residual potassium-40 is then subtracted from or added to the spectrum. Minimum detectable activity for Sr-90 then depends on the goodness of the background and candidate spectral matches, which allow spectra to meet form fitting criteria. 

The empirical basis for this photon-energy analysis is a mixed radionuclide standard. Strontium-90 analysis is then based on an Amersham S1Z.04 liquid standard sorbed onto cosmetic puffs to obtain reference density. This laboratory standard basis is then stepped down to lower activities in dry vegetable matter by reference to mulberry leaves obtained from N-Springs. This technique was piloted during joint samplings in the late 1980s. Incidentally, this instrument provides a gamma scan for each sample. Gamma emitters such as cobalt-60 and cesium-137 are identified. This laboratory-based, photon analysis system is linked to screening with a Ludlum 44-9 (pancake) detector which is relatively sensitive to beta emitters like Sr-90 in addition to gamma emitters like potassium-40. Samples can be selected in the field by careful counting under somewhat controlled conditions. A second count is obtained after a sample is dried (to 95°C). 

The presence of a beta emitter like Sr-90 is then indicated by an increased, elevated count on this survey instrument. An approximation is used to estimate the Sr-90 activity that will be counted on the spectrometer. Study validation involves requests of an agency to confirm particular sample results. This request has been declined. Submission of a sample to a previously used commercial lab has also been declined for this study. Therefore, falsifiability and accountability are maintained by making the analyzed material available to interested parties and by identifying to agencies the material and location of sample sources. To confirm Sr-90 in mulberry leaves at the H-Reactor Area shoreline, realism and consistency checks between data points and the published literature were also used. This validation provides an estimated confidence of at least 90% that the reported data quality is adequate for the intended purposes. This confidence has been considered in light of the public interests involved and found sufficient for the purposes. 

Conclusions and Recommendations 

1. Strontium-90 very probably co-migrates with hexavalent chromium into the Columbia Riverbed one or more fall chinook salmon spawning grounds in the Hanford Reach, at very localized activities above the drinking water standard of 8pCi/L. 

2. Strontium-90 is probably of greater ecological concern for natural spawning fall chinook salmon in the Hanford Reach than hexavalent chromium, for which remedial actions have been undertaken. 

3. Strontium-90 has not been previously identified at the Hanford salmon redds. 

4. Preliminary, appropriate mapping of intrusions of above-standard radionuclides into the riverbed of the Hanford Reach at fall chinook spawning grounds demands high priority.

References and Notes 

[1] D.D. Dauble and D.G. Watson, 1990, Spawning and Abundance of Fall Chinook Salmon (Oncorhynchus tshawytscha) in the Hanford Reach of the Columbia River, 1948-1988, PNL-7289, Pacific Northwest Laboratory, Richland WA. 

[2] S.J. Hope and R.E. Peterson, 1996, Chromium in River Substrate Pore Water and Adjacent Groundwater: 100-D/DR Area, Hanford Site, Washington, BHI-00778 (Rev.0), Bechtel Hanford, Inc., Richland WA, pp. 1-1 to 1-2. Strontium-90 Adjacent to Fall Chinook Salmon Redds at H-Reactor Area of the Hanford Reach of the Columbia River Norm Buske 1402 Third Avenue, Suite 1215 Seattle, WA 98101

[3] S.J. Hope and R.E. Peterson, 1996, Pore Water Chromium Concentration at 100-H Reactor Area Adjacent to Fall Chinook Salmon Spawning Habitat of the Hanford Reach, Columbia River, BHI-00345 (Rev.1), Bechtel Hanford, Inc., Richland WA. 

[4] Ref. [3], pp. 1-1 to 1-3. 

[5] Ref. [1], pp. 5.7 to 5.9. 

[6] Ref. [3]; this is Photo 2-2, p. 2-12. 

[7] Ref. [3]; this is Figure 2-2, p.2-7/8, with groundwater flow direction arrow drawn in. Groundwater flow direction is from Ref. [8], Figure A.4, p. A.13, and from technetium and uranium plume orientations in Ref. [8], Figures 5.7-1 and 5.7-2, pp. 5.146 to 5.147. 

[8] M.J. Hartman and P.E. Dresel (eds.), 1998, Hanford Site Groundwater Monitoring for Fiscal Year 1997, PNNL-11793, Pacific Northwest National Laboratory, Richland WA. 

[9] Ref. [8]; this is Figure 5.7-8, p. 5.151. 

[10] N. Buske and L. Josephson, 1986, Spring 1986 Data Report, Hanford Reach Project; now distributed by Nuclear-Weapons-Free America, 1528 West Sixth Avenue #3, Spokane WA 99204. 

[11] W.D. McCormack and J.M.V. Carlile, 1984, Investigation of Ground-Water Seepage from the Hanford Shoreline of the Columbia River, PNL-5289, Pacific Northwest Laboratory, Richland WA. N. Buske and L. Josephson, 1989, Water and Sediment Reconnaissance of the Hanford Shoreline, , Hanford Reach Project; now distributed by Nuclear-Weapons-Free America, 1528 West Sixth Avenue #3, Spokane WA 99204. R.E. Peterson and V.G. Johnson, 1992, Riverbank Seepage of Groundwater Along the 100 Areas Shoreline, Hanford Site, WHC-EP-0609, Westinghouse Hanford Company, Richland WA. 

[12] ___________, 1987, Subsurface Transport of Radionuclides in Shallow Deposits of the Hanford Nuclear Reservation, Washington--Review of Selected Previous Work and Suggestions for Further Study, USGS Open-File Report 87-222, Tacoma WA. 

[13] S.P. Van Verst, G.W. Patton, et al, 1998, Survey of Radiological Contaminants in the Near-Shore Environment at the Hanford Site 100-N Reactor Area, PNNL-11933, Pacific Northwest National Laboratory, Richland WA. This provides a reference list. 

[14] Ref. [8], Figure 5.5-1, p. 5.133. 

[15] R.L. Dirkes and R.W. Hanf (eds), 1998, Hanford Site Environmental Report for Calendar Year 1997, Pacific Northwest National Laboratory, Richland WA.

source: http://www.whistleblower.org/article.php?did=372&scid=124 18apr03