Thursday, January 7, 2010

Tritium in Groundwater and the RACER Database

The Department of Energy (DOE) has sponsored the creation of an online database and data analysis tool called Risk Analysis, Communication, Evaluation, and Reduction (RACER) located at This tool is being administered by the New Mexico Community Foundation (NMCF) and has been publicly available since early 2009. Currently, it is said that there are approximately 7 million pieces of environmental data accessible through RACER. Presumably, most of these data pertain to contamination left by the DOE's nuclear weapons program at Los Alamos National Laboratory (LANL).

RACER is also a tentative component of the DOE's newly refurbished Environmental Justice (EJ) Program. DOE says that the its EJ program aims to assist members of economically disadvantaged communities across the United States to participate in evironmental decision-making. This would be a decision-making related to the disposition of detritus from large DOE operations; in particular, to the accumulated legacy waste from its nuclear weapons program. But, whether or not RACER will prove to be useful in this regard, remains to be seen.

Quoting from a January, 2010 letter from the NMCF to RACER users:

"RACER is the first project ever in the United States where the public has full access and full transparency to environmental data at a Department of Energy site. It's giving the public a voice in a way that it never really had before."

Continuing from the same letter (but paraphrasing somewhat):

Do you have questions about what chemical or radionuclear contaminants may be present in the Northern New Mexico environment?

For example, is the soil on the Pajarito Plateau contaminated with chemical residues from high-explosives tests conducted at Los Alamos National Laboratory?

Has laboratory analysis confirmed the presence of radionuclides along your favorite hiking trail in Bandelier National Monument?

Have tritium levels in local groundwater ever exceeded national drinking water standards and, if so, where and when did this occur? (end of paraphrase)

Prior to the appearance of RACER, much of the groundwater data that it contains could be found in LANL's Water Quality Data Base (WQDB). However, the WQDB itself was withdrawn from public access in early 2009.

In my opinion, the extent to which RACER's data will act to "give the public a voice" is doubtful. Nevertheless, as an example of the kind of questions that could be raised by perusing this data, I've copied here a white-paper which I wrote in mid-2008, while a member of the DOE-sponsored Northern New Mexico Citizens Advisory Board. The paper was concerned with tritium activities in groundwater on the Pajarito Plateau, and was based entirely on data taken from the WQDB. No response to the questions raised in this paper has ever been offered by representatives of LANL, or of the DOE.



In an earlier communication(1), the first of a series, I presented a brief summary of the contents of LANL’s online Water Quality Data Base (WQDB). This was intended to serve as a non-technical introduction to the WQDB for members of the Northern New Mexico Citizens Advisory Board (NNMCAB). Although no feedback was evoked from NNMCAB members, reaction to this overview from LANL staff was quick, and in the course of a long face-to-face conversation I was presented with many critical LANL thoughts(2).

In this, the latest installment in my ongoing study of the WQDB, I react to some of these criticisms, while I present a WQDB based history of tritium concentrations measured all around the LANL site, beginning in the 1960’s, and extending to the present day. I continue to hope that my efforts will help to inform the future work of the NNMCAB.

Throughout the following, some paragraphs are repeats taken, with minor modifications, from my earlier paper(1).

Data tabulated in the WQDB ( is a record of the activities of specific radionuclides, at particular locations (wells or springs), usually in units of pCi/L (picocuries per liter). Uncertainties in measured values of activities are also usually tabulated, as well as the estimated(3) Minimum Detectable Amount (MDA). In many cases, the listed MDA is greater than the measured value of the activity, and its associated uncertainty. If so, then the measured activity is accompanied by a < sign, indicating that the measurement is thought to be of a “non-detect”(4). Often values of measured activities appear with a minus sign, indicating that the measured activity was less than the activity of a corresponding blank sample and, therefore, also of a non-detect(5).

Note that, generally, for activities recorded prior to 1995, no value at all for the MDA appears in the WQDB. In the following, I will not refer to data preceded by a < sign, or by a minus sign.

History of tritium activities

Keep in mind, while perusing the following, that the DOE sets the upper bound for acceptable values of tritium activity in drinking water at approximately 20,000 pCi/L.

Tritium, a beta emitter and essential to the weapons program, has been found to be widespread in groundwaters of the Pajarito Plateau. Levels in the alluvial wells of Los Alamos and Mortandad Canyons have been very high, until within the last ten years, when levels have subsided. For example, at LCAO-2 a value of 190000 pCi/L was measured on 8-13-70, but levels have decreased since then to a value of 64 pCi/L, returned on 1-15-08. At MCO-3 a level of 12000000 pCi/L was measured on 3-24-87, while the level was 5400 pCi/L on 7-12-04. Levels measured at drinking water wells have been moderately high until as recently as the 1980’s when they were found to be at 5900 pCi/L at PM-1 on 9-10-81, and 4200 pCi/L at PM-3 on 3-30-82. However, the measured levels at drinking water wells have now subsided to values of less than 1 pCi/L.

Levels measured at some wells of recent origin are also high; e.g., at the intermediate depth regional wells, such as R-6i, levels were 4300 pCi/L on 11-17-05, and 3800 pCi/L on 1-23-08. Levels measured at a few deep regional wells are a bit high. For example, a level of 195 pCi/L was returned from R-28 on 11-29-07, and a level of 28 pCi/L was returned from R-15 on 2-25-08.

Since 2001, levels returned from the Buckman wells have all been below 2 pCi/L. Prior to these recent measurements, there were only a very few measurements recorded. These occurred in 1973, and returned levels of approximately 100 pCi/L.

The highest recorded values of tritium activity are depicted in Fig. 1. Here, the history of detected tritium activity is displayed for the three sites, MCO-3, LAO-1, and PM-1. The first two of these sites are alluvial wells (shallow wells), one in Mortandad and one in Los Alamos Canyon, and the third site is a Los Alamos County drinking water well (deep well).

The similarity between these three histories is striking. The measured values of tritium activity were very low prior to 1970, jumped to high values, often suddenly, remained at a constant level for up to 2 decades and then, beginning around 1990, subsided slowly at a rate which was characteristic of the site.

In fact, this same behavior can be seen to have occurred at many other LANL locations. For example, Fig. 2 records the histories of tritium activities at the Mortandad Canyon alluvial well MCO-5, the Los Alamos County drinking water well PM-3, and the deep characterization well DT-10, located at TA-49. It seems that all these histories suggest a series of events in which some source of tritium contamination first appears around 1970, persists at a constant intensity until about 1990, and then subsides, either gradually or abruptly.

Looking back at Fig. 2, the decrease of the measured tritium activity with time, between 1987 and 2007, is very close to an exponential, with a resident half-life which I estimate to be about 7.6 years. Similarly, assuming an exponential decrease in tritium activities for the DT-10 and PM-1 histories too, I arrive at a value for the resident half-life in these two wells of about 12.3 years. Returning now to Fig. 1, I estimate the resident half-life of tritium in LAO-1 to be about 7.9 years.

Thus, it seems reasonable to conjecture that the deep wells PM-1 and DT-10, whose tritium resident half-lives I estimated to be 12.3 years, show time histories which reflect the natural decay rate of tritium (T1/2 = 12.3 yrs., for the beta decay of 3H), and suggest that the natural residence time for waters in the vicinity of the relevant well screens in these wells is much greater than 12 years. On the other hand, the shallow wells MCO-5 and LAO-1, have tritium resident half-lives which I estimated to be 7.6 to 7.9 years, suggesting that there is an exchange of waters in these wells with the surrounding area which takes place over a period of approximately 7 years.

The behavior of all of these histories of detected tritium activities is interesting, and has led to some prior well-informed speculation. In a August, 2005 paper(6), it was suggested by scientists working under contract to the DOE that this behavior could be traced to the contamination of groundwater by the Omega West reactor(7), formerly located at the head of Los Alamos Canyon. This reactor operated continuously from 1956 until 1992, when it was shut down for servicing, and soon thereafter decommissioned.

During the shut-down it was discovered that there was a leak of a significant size in a pipe buried underneath the reactor. This was a pipe which transported primary coolant water between the reactor and its external cooling tower. [Evidently, there was no provision made in the reactor design for a secondary coolant system, so that the primary coolant system could be completely contained within the reactor building, as is the case for all existing NRC licensed commercial power reactors.] At that time, it was said that it was not known for how long that the leak had existed.

[As is well-known, a fission reactor’s primary coolant should be completely isolated from the environment, since it contains a build-up of tritium produced by absorption of neutrons emitted by the 235U fuel. In cases where the fuel cladding has developed cracks or holes, highly radioactive fission products may also be released into the primary coolant. Among these is 99Tc.]

The radionuclide 99Tc, an electron(beta) emitter, is a common product of the fission of 235U, and does not occur naturally on the earth, except in the minutest of quantities. In fact, 99Tc is normally found in groundwater at activity levels as high as 1 pCi/L only in the immediate vicinity of a water cooled nuclear reactor. There, it can be ejected into the coolant during fission of the uranium fuel in the core, and may be released into the environment with coolant water effluents.

During the period 1956 to 1994 the Omega West reactor operated at the head of Los Alamos Canyon and occasionally deposited its wastes, containing 99Tc, directly into the canyon(8). Thus the presence of technetium in groundwater downstream of the now decommissioned Omega West reactor is a probable remnant of the operation of that reactor.

The high concentrations of 99Tc in the alluvial well LAUZ-1, in Los Alamos Canyon, and in the alluvial wells MCO-5, MCO-6, MCO-7, MCO-7.5, MT-1, MT-3-, and MT-4, in Mortandad Canyon, and in the intermediate wells MCOI-4, MCOI-5, and MCOI-6, in Mortandad Canyon, can be taken as evidence of contamination by effluents from that reactor. For example, 99Tc has been detected in groundwater taken from LAUZ-1 at levels of up to 38 pCi/L; in groundwater from MCO-7.5 at levels up to 23 pCi/L; and in groundwater from MCA-5 at levels up to16 pCi/L.

The regional wells R-22 and R-34 have shown the presence of 99Tc in groundwater taken from the aquifer underlying Mortandad Canyon, at levels of 5 pCi/L. However, there have been no detections of 99Tc at locations removed from the Pajarito Plateau. In particular, this radionuclide has not been detected in the Buckman wells.


Based upon the evidence presented, it is arguable that a major contribution to the past contamination of groundwater by tritium, beneath certain parts of the Pajarito Plateau, was made by a leak in the Omega West reactor primary coolant system, and that this source of contamination had been ongoing for as long as two decades(6).

[Oddly, during all this time, measurements of tritium activity in groundwater were also ongoing and evidence of the contamination of groundwater by tritium was persistent. It may seem difficult now to imagine how such data could not have been seen, at the time that it was being collected, as evidence for the existence of a major reactor leak. However, for as long as two decades, apparently nothing was done to put a stop to this potential hazard. Perhaps the hazard was judged to be of no great significance for the public health.

In this context, it is interesting to note that the Agency for Toxic Substances Disease Registry (ATSDR), an arm of the United States Public Health Service, issued a report(8) in April of 2005, in which they found no evidence of impacts to public health from LANL operations, over the period 1980 to 2001. However, no mention is made in this report of the very high tritium activities, recorded during the period 1980-1995, in the groundwater of alluvial wells located in Los Alamos and Mortandad Canyons. Concerning tritium, the ATSDR only reports, without comment, measurements of tritium activities in certain biota; e.g., 27,000 pCi/L in honey from Mortandad Canyon, sampled in 1980; 11,000 pCi/L in elk bone, sampled in 1994-5; and 17,000 pCi/L in “produce”, collected in 1999. The ATSDR is unalarmed by these values since the DOE sets a standard for the tritium contamination of biota at approximately ten times its standard for tritium in drinking water.]


(1) “Summary [with commentary] of “Radionuclides in Groundwater” from LANL’s “Water Quality Data Base”, K. LaGattuta, NNMCAB, 7-18-08.
(2) Meeting called at LANL by Lorrie Bonds-Lopez to inform me of my errors: in attendance, besides myself, were Ms. Lopez, Ardyth Simmons, Terry Morgan, and Mike McLaughlin (briefly present), Aug., 2008.
(3) As pointed out by LANL’s Terry Morgan, the Minimum Detectable Amount (MDA) or, Minimum Detectable Activity, is the result of a “calculation”. It seems to me, however, that this calculation is only an approximation to some exact theory of error, and must be problematic to some degree. Therefore, it is fair to refer to it as an “estimate”. Note that, prior to ~1995, Results appearing in the WQDB are unaccompanied by any value at all for the MDA.
(4) The level at which the recorded activity of a particular radionuclide is said to be a “detect” is somewhat arbitrary. In fact, at various times this level has been set at the MDA plus one standard deviation (SD); at other times, a “detect” has been assigned as the MDA plus two SD’s; and at still other times, a “detect” has been set at the MDA plus three SD’s. For cases in which a measured activity is close to the MDA, it is necessary to ask: Is this value, returned by some analytical laboratory, a real measure of the activity of a particular radionuclide, or a measurement error of some sort? The answer can best be couched in terms of probabilities. Generally, as a measured value approaches the MDA from above, the probability of its being a real value and not an artifact should approach zero. But, the probability of a measured value of activity at, or even below, the MDA, being a real measure of activity, with non-zero probability, depends upon one’s estimation of the reliability of the precise value of the MDA. (The SD is an accepted measure of the most probable variation from the mean found in a lengthy data series. For the data recorded in the WQDB, the SD could be taken as the tabulated Uncertainty.)
(5) If the measured activity of a sample (S), inside a container, is less than the apparent measured activity of the empty container (C), then a negative value appears in the Results column of the WQDB; i.e., since Result = S-C. The data errors accumulated over the course of a lengthy series of sample measurements are partly due to the variable influence of the individual sample containers upon the measurement process.
(6) In, “A Vadose Zone Flow & Transport Model for Los Alamos Canyon, Los Alamos, New Mexico”, by Bruce A. Robinson, etal, published online, August 26, 2005. In this numerical simulation of tritium transport beneath the Pajarito Plateau, the authors estimate the source concentration of tritium in the Omega West reactor’s primary coolant water to be ~18 x 106 pCi/L. They also describe the history of detected tritium activities from LAO-1, and from LAO-3, both alluvial wells in Los Alamos Canyon.
(7) “Lab completes Omega West reactor decommissioning”, LANL Public Affairs, July 30, 2003.
(8) In, “LANL Public Health Assessment”, ATSDR/US-HHS, April 26, 2005. It seems that the ATSDR undertook this study, in part, because of worries expressed to it by members of the general public, living around Los Alamos, about possible deleterious health effects arising from past LANL operations. It is interesting to note a passage appearing on p94 of this Assessment: “Three underground storage tanks were used to hold liquid radioactive wastes produced when the Omega West reactor was active. Usually, when the tanks were full, their contents were emptied and transported to the TA-50 Radioactive Liquid Waste Treatment Facility. Occasionally, when the tanks were full, liquid waste water was discharged directly into Los Alamos Canyon.”

Fig. 1

Fig. 2

No comments: