Synthesis/Regeneration 11   (Fall 1996)

Citizens' Radiological Monitoring Network Detects Radioactive Cloud Downwind of Seabrook Station

by Sam Miller, New Hampshire/Great Bay Region Greens



The C-10 Radiological Monitoring Network (C-10/RMN) has been monitoring airborne radioactive emissions from the Seabrook nuclear plant since 1990. We have about 25 stations inside Seabrook's 10-Mile Emergency Planning Zone in southern New Hampshire and northeastern Massachusetts. A typical station, mounted on the home of a volunteer, consists of a beta/gamma detector, a gamma-only detector, and a weather station. These instruments feed minute-to-minute data into a computer inside the house, which stores it for later retrieval. Eventually the data is collected at our office in Newburyport, where we analyze it.

Monitoring and Modeling

Monitoring is the work carried out by the C-10/RMN and other citizens' radiological monitoring networks. It is performed by several people with radiation detectors at many locations away (i.e., potentially downwind) from a nuclear power plant. Once the network's individual sensor records are collected in one location, a person familiar with streamline analysis can plot the data on maps of the area, and trace a radioactive "plume" back to its origins. Such analyses are irreplaceable tools for activists concerned with holding the correct parties responsible for radioactive trespasses on their communities.

In modeling a plume, one uses meteorological and radiological data from a single geographical point-the nuclear power plant itself-to predict the shape a plume will take once it escapes from the plant. The model incorporates information about the magnitude of the radioactive release, the weather at the time of the release, and the architectural features of the plant, and makes predictions about radioactivity levels at all points away from the plant. Modeling is a powerful tool for gaining insighte into the nature of a plume detected by a citizens' network.

In the early morning of November 29, 1995, three C-10/RMN stations located southwest of Seabrook indicated the passage of a radioactive cloud. The beta/gamma detectors jumped as high as 8 times normal background levels and remained there for several hours, while the gamma-only detectors went as high as 15 times normal background. Our weather stations indicated that the wind was blowing steadily from the northeast throughout the period, so we suspected that the Seabrook Station was the source.

We occasionally register elevated readings (though never before as high as those on November 29) at one or more C-10/RMN stations, and follow up by checking with plant spokespeople or the Nuclear Regulatory Commission's (NRC's) on-site inspector. Usually we have an answer in 2 or 3 days, and it often turns out that the high readings coincide with various "evolutions" within the reactor complex. This time our inquiries went unanswered for weeks. Only after we announced that we were going to the press did Seabrook's spokespeople get back to us.

Their story was that the plant had indeed been venting radioactive material at the time of our high readings. They claimed that the material was tritium-a form of heavy hydrogen-which had been vented at rates so low that our sensors should not have picked it up. Furthermore, they acknowledged that the plant's Wide Range Gas Monitor had been inoperative since about 30 minutes before the venting began, and that "periodic sampling" of the outflow was performed as a back-up.

There were two problems with this. The first was with the specific radioisotope involved. Tritium does not emit gamma photons, but the highest C-10/RMN sensor response was from our gamma-only detectors, indicating that the cloud could not have consisted of tritium alone. The second problem was with the rate at which Seabrook admitted venting. We have our own computer model for simulating the dispersion of radioactive clouds. It's based on the same mathematics and references as the models used by the nuclear industry and the NRC, and was developed with the assistance of several scientists. Using this computer model, which accounts for factors of weather and plant construction, we found that the rate at which Seabrook admitted venting could not have caused a cloud of the magnitude we observed. Instead, we found that a release rate about 50,000 times greater than Seabrook admitted was required to reproduce our cloud.

It's possible that something is wrong with the computer model, but the weather conditions were such that it should have worked quite well. If the computer model failed under these nearly ideal conditions, then nobody should be using it. That means that the nuclear industry and the NRC should stop using it to predict radiation doses to the public.

We went ahead with our press conference in mid-March. Several reporters told us that the night before the conference, Seabrook's public relations people called them and put their own spin on events, denying that the plant was responsible for our high radiation readings. The following day several newspapers in the region ran the story. The Portsmouth Herald used it as their lead. A number of radio stations, including New Hampshire Public Radio, mentioned it in their headlines. Follow-up stories, editorials, and political cartoons, most of them favorable to our side, continued for several weeks. The Boston Sunday Herald picked up the story about a month later as their lead, with a full two-page spread inside. Our field director was interviewed on cable television.

The day following our press conference, Seabrook Station's lobbyist was in the Massachusetts state capitol, trying to get our funding cut off or reduced. (We receive an annual grant from the state to assist with equipment purchases.) Seabrook's PR experts went to work trying to discredit us in the press, attacking our data and our instrumentation, and offering several "alternative" explanations for the high readings. One of their suggestions was that precipitation might have brought down fallout from the atmospheric bomb tests. We checked this out, and found that Seabrook's own weather records indicated that snowfall had not even begun until two hours after the radiation readings shot up. Nothing happening at the plant could possibly have resulted in such high radiation levels, they said, but they were just as determined as we were to get to the bottom of it.

After learning of our intention to go to the media, the Seabrook managers began desperately trying to arrange a meeting with us. Several of us received a number of phone calls at home-an entirely new phenomenon. (Prior to this, I couldn't even get responses to certified mail.) We agreed to meet with them as soon as possible, and a week after the press conference, eight C-10/RMN activists went to the plant for a meeting. When we arrived, we were greeted by 3 PR people, 2 plant engineers, 2 environmental monitoring people, a health physicist, a meteorologist, and several other plant employees.

We were given an elaborate tour, including the vault containing the huge turbine, the control room, the on-site radiation monitors, the health physics area, and the computer program they use to calculate off-site radiation doses. Following the tour, we gathered in a conference room, where we were given an overview of the plant's off-site monitoring program and fed another assertion that the plant could not have caused our high readings. According to plant spokespersons, it was down at the time of the readings, and was not generating any radioactive gases. They had gone back and checked their own off-site monitoring records at the time and found nothing abnormal.

The Seabrook off-site monitoring program basically consists of thermoluminescent devices [TLDs] scattered around on phone poles, and eight gas samplers. The TLDs are taken down and checked every three months, and the gas samplers are checked once a week. But TLDs have been shown to be highly insensitive to noble gases-the most likely composition of the cloud we observed. And since plant employees collect TLDs only once every three months, they average a short-duration "spike" over the whole period, making it disappear. A relatively narrow plume could easily move between 2 of the 8 gas samplers without registering in either. The computer model predicted this behavior, and we observed it with our network.

Then it was our turn to speak. We discussed the two problems with their story: the lack of gamma emissions from tritium and the vast disparity between the rate of release they admitted and the cloud we observed. An odd collection of facial expressions greeted the suggestion that their numbers were 50,000 times too small: the poker faces of the PR specialists, the surprise of the engineers, the outrage of the environmental monitoring people. They attacked the validity of our sensors and our computer model. We replied that the sensors were factory calibrated in a noble gas atmosphere and independently tested at a university, and that the computer model was mathematically identical to the one they were using. At times the exchange was heated.

The meeting went on for three hours. In the end we agreed that we would try to exchange information more openly. One of their PR specialists agreed to be the point man for communications between the C-10/RMN and Seabrook. We all shook hands and went home.

The next day I spoke with a nuclear engineer, formerly of Northeast Utilities (which owns Seabrook), who is now a whistleblower. I asked him if it was possible for Seabrook to be the source even if it wasn't operating at the time. Yes, he replied, the plant could still be the source. Radioactive gases created while the plant is operating are collected into a delay system and allowed to decay. The gases may be delayed in this system for a period of several weeks, but if the valves on the system are opened, a release rate of "curies per minute" can result-more than enough to account for our plume.

I wrote to our "point man" at Seabrook with this information. A month later he replied with another denial, while not directly responding to the whistleblower's suggestion. Then he said he would no longer communicate with me. From now on we would have to communicate with Seabrook through the state of Massachusetts.

Since then we've done some more research, and found that Seabrook's waste gas system can't simply be "opened up and vented" as the whistleblower had suggested. (This is possible on older plants, such as the Millstone reactors.) But the Massachusetts state nuclear engineer has become interested in the issue, and is not satisfied with the responses he's received from Seabrook or the NRC. As a result, early this summer, the NRC will conduct an entirely new inquiry into the events of November 29, 1995. We are considering attempting to place a volunteer observer on the panel.

These are the highest radiation readings our network has ever observed, and the wind was blowing from the plant to the sensors that registered the high readings. The plant's spokespeople, backed up by the NRC, continue to insist that their reactor could not have been the cause, leaving us without a known source for the cloud. But the C-10/RMN detects smaller radiation spikes once or twice a month. The need for citizen involvement in radiological monitoring is clear.


Sam Miller is a member of the New Hampshire/Great Bay Region Greens. He is an applied meteorologist, holds a degree in physics, and has been a C-10 volunteer since 1991.

This article as marked up by Sam Miller: http://pubpages.unh.edu/~stm/Why_Monitor.html

Where To Go For More Information.


Synthesis/Regeneration home page  | Synthesis/Regeneration 11 Contents