Purpose/type: nuclear center for weapons production

Location: a 180-ha area at Salives (Cte d’Or), 30 km northwest of Dijon

Operator: Direction des applications militaires du CEA

Period of operation: since 1957

Nuclear materials handled: uranium, plutonium, tritium, lithium, deuterium

In particular: produces the nuclear portions of nuclear weapons in series

When it opened, the center was only an annex of Bruyères-le-Châtel . The CEA acquired an area of 588 ha, of which 413 ha were wooded, but the center proper occupies only 180 ha [CDRPC 94].

As the result of a reorganization of the Dam decided upon in 1996, Valduc will group for the Dam all competencies connected with nuclear materials. It will conduct research, fabricate weapons, and treat radioactive materials [Bataille 97]. The CEA has transferred from Bruyères-le-Châtel to Valduc activities in these areas and the CEA employees who carried them out.

For studies concerning phenomena under radiation, Valduc has accelerators, including Lancelot. The center also has means of gamma simulation, transferred from Cesta around 1998, and two sources of neutrons, Prospero and Caliban, put into service in 1972 [DAM x.72].  Caliban, at least, was still in service in 2003.

A criticality station was put into service in 1963 [défi iii.96]. It operated/operates Maracas (Machine de rapprochement pour la criticité d’assemblages solides) and the reactors Mirène (Mini-réacteur pour neutronographie ) and Silène, and Appareillage B (Equipment B).  

Maracas, put into operation in 1982, was not operating in 2000. 

Mirène (a shut-down homogeneous reactor) operated from 1975 to 1982. Its fuel was enriched uranium (1.2 kg in 10 l of solution); its thermal power was not used. 

Silène (Source d’irradiation à libre évolution neutronique) which is operated today by IPSN, entered into service in 1974. Its fuel is a solution of 40 l of uranyl nitrate containing uranium enriched to 93% [LeMo 23.vi.93]. Its power in continuous operation can reach 1 kW. In pulse operation, the reactor reaches 1000 MW [AIEA 95].  

Appareillage B, a subcritical assembly, was put into operation in 1963 and was renovated in 1997.  It "is composed of a network of fuel rods partially under water" and "makes possible the determination of risks involved in the manipulation of new fissile milieux, like Mox [and] high burnup fuel" [Ener 15.xi.99].

RACHEL

The fast neutron reactor Rachel, put into service in 1961, was probably the first reactor used for criticality experiments. The core had two hemispheres of plutonium metal surrounded by a natural uranium reflector. Its biological protection was assured by distance, and the control installations were located 120 meters form the reactor [Bourgeois 62]. The annual reports of the CEA do not mention Rachel after 1971.

ECHO

A two-stage launcher (one stage with powder and one with compression) which sends a projectile (reaching speeds greater than 3000 m per second) to a radioactive target enclosed in a small, tight target carrier. It is probable that the target carrier contains plutonium 242.

Cylindrical tank in which the DAM generally executes cold firings, which are firings in a vacuum with small quantities of explosives associated with "materials carrying risks of a radioactive and chemical type." The substance used at Guillaume Tell is often, if not always, plutonium, probably plutonium 242 [Choc ix.93; Science vii.96].

A 1998 report from the Dam on contamination at several of its sites, names, among the principal installations at Valduc that present a potential risk of radiological or chemical contamination, a building for the production of parts from plutonium, installations for the production of parts from uranium, a building for the production of ensembles containing tritium, and a building for the storage of fissile materials, [HC 98]. Specific facilities include workshops for making alloys, for melting and for machining, and for assembling and checking parts made from depleted uranium, enriched uranium, and plutonium; and also equipment allowing the manipulation, for fusion weapons, of tritium, deuterium, lithium, and their compounds, including lithium hydride (LiH) and lithium deutero-tritiide (DTLi).

The fabrication of parts from uranium and plutonium uses the same technologies as for other metals, but in glove boxes. Work on light fusion materials is carried out in lines that take care of the purification of components, the elaboration of components, and the making of weapons parts [DAM vii.76]. Miramas furnishes the LiH, but it is possible that Valduc produces the DTLi.

The assembly of nuclear warheads has never been carried out at Valduc, but from 1980 to 1993 the air force operated a center for assembling nuclear warheads for the army and air force on a 40-ha tract that had been part of the site and that the defense minister acquired from the CEA. The assembly of ASMP and Hadès missiles was carried out by the air force on that tract [CDRPC 94].  Today that parcel is the site of the Centre de stockage militare de Valduc (CVSM, Valduc Center of military storage).  While awaiting dismantling, nuclear subassembles are stored there [Obsan #2, iii.00].

The maintenance of nuclear warheads in service in the armed forces is listed among the activities at Valduc. The majority of the warheads contain tritium and, because of the short half life of tritium, the tritium in them must be replenished periodically. It is probable that filling with tritium the containers to be placed in warheads occurs at Valduc [Barrillot 01].

According to Christian Bataille, the plutonium in nuclear weapons also has to be maintained. Warheads include plutonium 241, which is shorter lived than the fissile plutonium 239 and which decays into americium 241. Americium 241, a neutron absorber, reduces the efficiency of weapons. Thus, in order to maintain arms at their initial power, americium 241 has to be periodically removed. Its elimination is accomplished by the electrochemical process described in "Extraction of americium" under "Dismantling of nuclear subassemblies" below.

When a nuclear warhead reaches the limit of its life, it is disassembled and nuclear components are sent back to Valduc. The center recovers and recycles the materials. We do not have details concerning the recycling of the uranium and the tritium, but the treatment and the manipulation of the plutonium have been described by Philippe Guiberteau of the Section retraitement plutonium in Le Bulletin Dam, October 1995. The following paragraphs on the treatment of plutonium are based on this article.

The nuclear subassembly is opened by machining on a lathe isolated in a glove box to recover the plutonium shell. The recovered plutonium is melted in an induction furnace to form masses weighing several kg each. The plutonium then undergoes treatments to make it reusable.

To make the manipulation of plutonium in classic glove boxes possible, the level of americium must be reduced to less than 0.1%. The first stage is an electrochemical process. The plutonium is melted at 800° C in a ceramic crucible in the presence of alkaline salts. It is partially oxidized to Pu3+, and the ions react with the americium to produce Am3+. At the time of cooling, 90% of the americium remains locked in the salts. The plutonium metal is shaped by fusion.

The plutonium from which the americium has been removed is purified by an electrorefining process, likewise with melted salts. It is melted at 750° C in the presence of alkaline salts and oxidized by electrolysis into Pu3+ on the anode, then reduced in Pu on the cathode. The metallic impurities are not oxidized. A round of pure plutonium is obtained in the cathode compartment after cooling. A block of directly usable metal is then obtained by melting the plutonium.

Dismantling of subassembles and the purification of plutonium, like fabrication operations, generate reprocessable residues. Valduc treats these products and those from Bruyères-le-Châtel on the site [CEAD 97]. During a series of operations—dissolving, precipitation, elaboration of the metal—the plutonium is recovered to be reused.

III.B.1 GASEOUS EFFLUENTS [see below]

III.B.2 LIQUID EFFLUENTS

Effluents "slightly" contaminated with transuranics are sent to Andra. According to Guiberteau in 1995, effluents more heavily laden with americium are concentrated by precipitation with sulfur. The supernatant is sent to Andra, and the concentrates are vitrified. The CEA’s waste management department reports that all liquid wastes from the treatment of plutonium with activity over 4500 Bq/cm3 are vitrified [CEAD 98].

At least a portion of the highly radioactive effluents themselves or concentrates resulting from them have been dispatched to the vitrification workshop at Marcoule. Bataille wrote in 1997 that 2.1 m3 of effluents went to Marcoule in 1996 and that 3.5 m3 of very highly active effluents were still stored at Valduc.  According to him, the storage capacity at Valduc for highly radioactive wastes should suffice into 2003  [Bataille 97]. The CEA states that highly radioactive wastes in general are vitrified at Marcoule [CEAD 98].  

Contaminated heavy oils were sent to Cadarache "to be burned in specialized installations." Initially Valduc had a stock of 11m3, but 6 m3 remained to be burned in 1997 [Bataille 97].  According to Andra, 36 m3 of oils and soils contaminated with alpha emitters were stored at Valduc in 1999.  Their final disposition was yet to be decided upon [Andra 00].

III.B.3 SOLID PLUTONIUM WASTES

III.B.3.a Salts laden with americium.

III.B.3.b Other wastes too heavily contaminated with alpha emitters to be sent to Andra

In 1997, according to Bataille, the site stored the following additional solid wastes resulting from the recycling of weapons: 45 m3 of stable wastes from reprocessing, 82 m3 of wastes coated in concrete, 3.2 m3 of wastes coated in asphalt (both these methods of packaging alpha waste are no longer used), and 2.5 m3 of wastes resulting from operation TRIRAD (see below).

 Andra in 2000 said that Valduc stores the following three categories of alpha waste: 530 drums (16 TBq) of highly radioactive, long-lived wastes, coated in concrete or asphalt; 478 drums (24 5Bq) of wastes awaiting treatment; and miscellaneous wastes (3 TBq).  The last two categories include in part wastes destined for Andra and in part highly radioactive, long-lived wastes.  We do not know the precise relationship of the items in Bataille’s list to those in Andra’s inventory..

Bataille notes that from the information given to him by the Dam, he is unable to determine the exact quantities of plutonium left in the wastes [Bataille 97].

III.B.3.c Wastes sent to Andra

Bataille presents figures representing the dispatch of wastes from CEA-Dam in general to Andra. A large portion of those wastes would have come from Valduc. Since 1983, Bataille says, the Dam has sent a total of 5000 m3, initially to CSM, then to Soulaines. The quantity of plutonium that has accompanied these wastes is estimated to be about 9 kg (27 TBq, 750 Ci), of which 2.5 kg (8 TBq, 216 Ci) since 1985.

III.B.3.d Wastes sent to Cadarache

Cadarache, Bataille says, has received from other CEA sites since 1985, 500 m3 of waste packaged in drums. This waste, which could not meet the strict standards established in 1984 for Andra’s sites, contains an estimated 7.4 kg of plutonium (22.7 TBq, 614 Ci). Again, much of the waste would have originated at Valduc. Bataille points out that the estimated total of 16.3 kg of plutonium evacuated from CEA centers is not a negligible quantity.  The 2000 Inventory of Andra does not indicate that wastes from Valduc have been sent to Cadarache.

III.B.3.e Volume Reduction—Trirad

In order to reduce the volume of waste that must be disposed of, the CEA has constructed at Valduc an incinerator for solid, burnable wastes, too contaminated with alpha emitters to be sent to Andra. It will use the Iris process, developed by the CEA at Marcoule. The material burned will contain (by weight) 50% PVC, 35% latex or neoprene, 10% cellulose, and 5% polyethylene.

The authorization to begin use of radioactive materials was received in January 1999; and the facility was to begin incinerating radioactive waste in the first half of the year. Nominal capacity will be 7 kg/h; the installation is designed to treat 26 t of waste per year. It is designed to operate continuously.  The incinerator is located in a new three-story building.

After shredding and extraction of any pieces of metal, the waste will undergo the following treatment:

--pyrolysis at 550° Centigrade in a rotating furnace under a reducing atmosphere to produce pitch, high in carbon;

--calcination of the pitch at 900° C again in a rotating furnace but in an oxygen-rich atmosphere to obtain ash and gaseous effluents;

The off-gas will be cooled from1100° to 250°C; filtered in an electrostatic filter to remove, according to the CEA, "more than 99.8%" of the fines, chiefly volatile ash and zinc chloride; passed through "very high efficiency" filters; and finally neutralized by means of a new procedure using sponge lime at about 130°C.

In the course of incinerating 7 kg of waste, the incinerator will produce 0.2 kg of ashes and about 0.1 kg of fines. In these operating conditions, the incinerator would produce about 600 to 700 kg of ashes and 300 to 350 kg of dusts per year.  The CEA envisages long-term storage for the ashes; but it could recover them for treatment with Ag3 to capture the plutonium, or packaged in view of storage.  It would be possible to vitrify the ashes.  The future of the fines depends on their radioactivity, that is to say on the quantity of plutonium oxide entrained in the gaseous flow. Vitrification of the fines after packaging in a matrix would be possible [CEAD 99].  

The authorization for operation with radiaoctive material was delivered by the Haut Commissaire in January 1999, and three incineration campaigns took place that year.  "Difficulties appeared during the third campaign, leading notably to blockages at the level of the pyrolysis furnace."  An action plan for resolving the probem was put into place.  Three campaigns were planned for 2000 [CEAD 98].

III.B.3.g Volume reduction—dismantling

Valduc dismantles contaminated, bulky items of equipment and glove boxes in a "salle de casse" (room for breaking up), which centers in a dismantling compartment where operations are guided by remote control. The "salle de casse" has broken up more than 1000 m3 of voluminous objects since it began operation in 1993 [CEAD 98].

The wastes coming from the fabrication of parts in enriched uranium are treated at Valduc. André Malfondet, of the Valduc SPR, described the treatment of the solid wastes in 1988:

--turnings and masses are reinjected into the fabrication process;

--molds and crucibles of graphite, used in the foundry are treated "by grilling at 500° C under a current of air in such a way that they can be sent in 100 l drums" to Andra. The powder from the grilling, uranium oxide, is sent to Pierrelatte for recovery of the uranium. (100 kg/yr).

--wastes proper are compacted and sent to Andra [provided they meet Andra’s standards];

--sandblasting "allows us to put back into ordinary use an important volume of metallic wastes" [Compte 88].

Solid wastes contaminated with uranium may also be treated by Trirad and, presumably, when appropriate, by the incinerator after it begins operation.

(Wastes from the fabrication of weapons parts in natural uranium and depleted uranium entered into the circuit of the Dam that we outline under Bruyères-le-Châtel and under the Centre de Vaujours—Moronvilliers. The role of Bruyères-le-Châtel in the circuit may now be played by Valduc.)

III.D.1 GASEOUS EFFLUENTS

In 1992 Valduc oxidized tritium gas by catalytic reaction to produce tritiated water. The water was trapped on a molecular sieve that was periodically regenerated. The operation was expensive and, at Valduc, involved large quantities of tritium [Creusefond 92]. The CEA waste management department more recently alludes to equipment used to remove tritium from gas that is in containment structures, before the gas itself is extracted [CEAD 99].

III.D.2 LIQUID EFFLUENTS

At Valduc, tritiated water comes from procedures during which gaseous tritium enters into contact with oxygen, as well as from procedures for treating gaseous and solid wastes. Creusefond in 1992 wrote that tritiated water, containing up to 106 Ci per liter, is frozen and stored at -20° C to reduce its aggressiveness towards containers. Bataille stated in 1997, however, that tritiated water is stored in polyethylene bottles inside the production buildings. He described the water produced by the treatment of the solid waste as containing 10 to 500 Ci per liter, and said that the stock in 1997 amounted to 800 l [Bataille 97].

III.D.3 SOLID WASTES CONTAMINATED WITH TRITIUM

Solid wastes that have absorbed tritium are produced during the operation and dismantling of laboratories and workshops. In 1999 the CEA's waste management department the installations in service at that time for treating solid wastes were essentially only two in number. Bataille apparently described the same two, when speaking in 1997 of wastes with a high tritium content:

--equipment designed to melt metals that give off much tritium gas. They are treated in a furnace under a vacuum, which produces metal ingots with a low level of activity and a low rate of degassing;

--equipment to treat organic solid wastes with dry steam to reduce their activity. The wastes are then compacted and packaged.

Both the above methods produce tritiated water, which, as Bataille states, illustrates "the paradox of the management of tritiated wates," since the water is almost as difficult to manage as the original waste [Bataille 97]. The solid products of the treatment as well as the water are stored at Valduc.

A CEA/Valduc press dossier, dated March 28, 2000, mentions a third method.  It concerns "mineral sponges":  "high temperature treatment," packing in drums, then storage at Valduc.  Doubtless, this method also produces tritiated water.

Some solid wastes are stored without prior treatment. The three distinct categories are

--technological wastes, packaged in drums, with a relatively high degassing rate of between 1.85 and 55 MBq per drum. About 580 m3 of these wastes were stored at Valduc in 3032 drums in 1997;

--technological wastes, packaged in drums, with a "very low" degassing rate of 1.85 MBq per drum. About 550 m3 of these wastes were stored in 2915  drums in 1997 at Valduc;

--scrap iron with a "very low" rate of degassing and a surface activity of between 3.7 and 37 Bq per cm2.  About 220 t were stored in 1999.

Oil and mercury contaminated with tritium are also stored at Valduc [Bataille 97].

The tritiated solid wastes are in three locations:

--building 055, a hangar built on a slab and furnished with air extractors on top of chimneys. This building, which in 1997 contained 3200 drums is responsible for 10% of the tritium emissions from Valduc, according to the Dam;

--building 058, a building on a concrete slab, but without artificial means of ventilation and thus with a lower level of emissions. In 1997 it also contained 3200 drums;

--scrap metal on a concrete-covered area, 50 tons in 1997 [Bataille 97].

The facilities at Valduc for storing tritiated waste contain such waste from the various centers of the Dam (85%) and from other CEA centers (15%). The waste cannot be sent to Andra for final disposal, because it is not possible to characterize thoroughly a package that contains heterogeneous items with varying levels of tritium. The rate of degassing can be calculated but not the mass activity. To meet Andra’s standards, both must be known [CEAD 98]. As the Dam recognizes, storage of the waste at Valduc is not a final solution. The three storage units at Valduc have a total capacity of 1760 m3. In 1999 they contained l507 m3 of tritiated solid waste. The production rate is about 50 m3/yr on average. Valduc was in 1997 considering constructing a new storage building, and, at the request of the Haut Commissaire à l’Energie Atomique, a working group on "The future of tritiated waste" had been put in place [Bataille 97].

Between 1968 and 1975, drums of waste contaminated with tritium were burned in the open air at the old Chatellenot farm, located on 40 ha of land that the CEA turned over to the defense ministry in 1980 for weapons assembly [CDRPC 94]. According to recent reports from the CEA, burning took place on 43 occasions and involved 335 m3 of various wastes [Bataille 97]. During the same period approximately 30 tons of scrap iron contaminated with tritium were stored in an old cistern on the farm. At the insistence of the unions, the scrap was removed in 1981, after it had contaminated the underground water [see CDRPC 94]. A 1998 report drawn up at the request of the Haut-Commissaire à l’Energie Atomique on the contamination of the site discusses the burning but fails to mention the metal [HC 98].

We have very little information on wastes consisting of lithium and its compounds, but documents from the late seventies indicate that at that period these wastes were a major problem. The Dam did not know how to destroy LiH completely, nor what to do with the LiH or with the residues from its treatment, including lithine (LiOH) at the center [see CDRPC 94]. A 1998 CEA document reportedly mentions a deposit of LiH at the site [HC 98].

IV.A.1 AUTHORIZATION

A decree of 3 May 1995 states that the annual radioactivity of gaseous effluents from all buildings must not exceed a total of 1759 TBq (50 kilocuries). Specific limits are :

Tritium: 1850 TBq (50 Kilocuries)

Gasses other than tritium: 40 TBq (1 kilocurie)

Gaseous halogens and aerosols: 750 MBq (20 millicuries)

Alpha emitters: 75 MBq (2 millicuries) [Bataille 97].

Alpha emitters in aerosols: said to be below the limit of detection and thus estimated to be less than 440 kBq/year (10 µCi/yr)

Fission products (rare gases, iodine, cesium, strontium are susceptible to being released during an accident): "insignificant" quantities of beta emitters after passage of the gas through active carbon and HEPA filters; xenon 133 and xenon 135 about 1 TBq/yr (27 Ci/yr) [HC 98].

IV.A.3 PAST RELEASES

IV.A.3.a Tritium:

The maximum annual release of tritium was attained in 1975 with 21,000 TBq (568,000 Ci). Since, annual tritium releases have decreased and by 1997 reached 310 TBq (8400 Ci).

Accidental releases of tritium have occurred repeatedly. (Statistics on annual releases apparently do not include accidental releases.) The years reported to have the highest total accidental releases are 1970: 203,700 Ci; 1971 68,000 Ci in two releases; 1978 133,700 Ci in five releases. In 1994, two releases totaling 7290 Ci occurred. The CEA reports no accidental releases for 1992, 1993, and 1995. We do not have later figures [HC 98].

The maximum level of tritium in the plume in the immediate proximity of the burn site at the Chatellenot farm (see above) was 1 Mbq/m3 (0.27 mCi/m3), according to a recent CEA study [HC 98].

IV.A.3.b Fission products:

In 1971 37 TBq (1000 Ci) of fission products (rare gases) were released during criticality experiments conducted by IPSN [HC 98]. This accident was cited at a meeting in 1998. Others involving radioactive gaseous effluents other than tritium may have occurred.

Valduc is not located near a river or other large body of water. The authorization of May 3, 1995, for nuclear rejects therefore does not permit any release of radioactive liquid effluents. "Since January 2, 1996, no effluent that comes from the nuclear zones or has passed through them is released into the environment" [HC 98].

IV.B.1 PRESENT MANAGEMENT

Used water is piped to a conventional biological purification station. In 1996 this station was reconstructed and equipped with five storage tanks with a total capacity of 2500 m3. Effluents are tested and if they meet drinking water standards, released into the valley, the Combe au Tilleul (apparently once dry but now containing lagoons) below the station. Since 1987, sludges from the purification stations at Valduc and also at Bruyères-le-Châtel are stored at Valduc in a tight building dedicated to that purpose [HC 98; CEAD 98].

Effluents susceptible of being radioactive and with an activity under 4500 Bq/cm3 (500 m3/yr as of 1997) are sent in tank trucks to treatment installations where they are subject to precipitation, distillation, ultrafiltration, and finally evaporation. The solids remaining after evaporation are packaged in cement and sent to Andra [Bataille 97]. The CEA has been testing a new treatment method using extractive molecules that is proving to be particularly effective in decontaminating uranium solutions.  A pilot installation has demonstrated the process [CEAD 99]. (On the treatment of highly radioactive liquids, see Liquid effluents under Plutonium Wastes above.)

IV.B.2 PAST MANAGEMENT

Used water from the non-nuclear portion of the site (the lower zone) was treated until about 1994, at least in part, in a conventional biological purification station and then released into a pond in the Combe de Noirveau. Used water from the nuclear portion of the site (the upper zone) was sent to the treatment center near the Combe au Tilleul and, after treatment, released into the Combe au Tilleul.

Effluents susceptible of being radioactive were collected separately from the used water, the CEA reports. They were sent to a station for specialized treatment and then to "the biological treatment station to the south of the center," or went directly through pipes to that same purification station. The document describing the water treatment seems to say that the radiological effluents arrived at the same station as the used water from the site’s upper zone—near the Combe au Tilleul. After treatment, the effluents likely to be radioactive were released, apparently in the Combe au Tilleul, in accordance with the norms fixed by SCPRI and later OPRI.

The CEA/Dam has estimated the cumulative radioactivity of effluents released into the Combe au Tilleul before 1994 when radioactive releases ceased. Releases of radionuclides totaled: cesium 137: 3 GBq (0.1 Ci)—the cesium accompanied by a little strontium 90 resulted essentially from criticality tests conducted by IPSN; uranium 235: 0.74 GTBq (0.02 Ci); americium 241: 0.18 GBq (0.005 Ci); plutonium 239:0.18 GBq (0.005 Ci).

At the request of the High Commissioner, the Combe au Tilleul was rehabilitated in 1995 and 1996. At that time 7800 m3 of earth were dug up and placed in the structure at Valduc that holds evaporation sludges. The earth that was dug up contained the following volumetric activity: plutonium 239 and americium 241: 0.03 Bq/g; cesium 137: 0.3 Bq/g; uranium 235: 0.04 Bq/g. The earth remaining in the Combe has a radioactivity today of about 0.05 Bq/g in plutonium and in americium and less than 1 Bq/g in cesium 137 [HC 98].

Sludges from the purification stations, which had an alpha activity of a few bq/g were apparently spread on the lawns and inert dumps, according to the CEA. Then, for a while, they were sent to Andra’s La Manche site. In 1987, 74 t of sludges produced between 1982 and 1987 were put into a class 1 dump at Pontailler-sur-Saône [HC 98]. Andra reports that the sludge at Pontailler-sur-Saône contains less than 10 Bq/g of uranium and other transuranics [Andra 99].

IV.C.1 ATMOSPHERIC TRITIUM

The "average annual volumetric value for tritium" (gas and water vapor) above the site is on the order of some 10 Bq/m3. Off site, volumetric radioactivity is only several Bq/m3.

IV.C.2 RAINWATER

The volumetric radioactivity of rainwater falling on the site is on average on the order of 1000 Bq/l. Radioactivity of rainwater is higher near storage facilities for tritium. Within a radius of 6 km around the site, radioactivity is several dozens of Bq/l, up to 100 Bq/l.

IV.C.3 RIVER WATER

In 1997 the volumetric activity in tritium in the ruisseau de Noirvau was 500 Bq/l and, in la Douix, below its confluence with the Noirvau, 100 Bq/l. At spring R14, two km south of the Combe de Tilleul and a primary outlet for water absorbed by the Combe, activity in tritium was 600 Bq/l around 1997. At sources of the Seine, 15 km southwest of the Center, volumetric activity in tritium was 7 Bq/l in 1997, down from 63 Bq/l in 1984.

IV.C.4 UNDERGROUND WATER

The hydrogeology of the site is typical of limestone regions, where runoff moves down through cracks in the rock and circulates quickly within a complex network of underground channels, to appear again at the surface in springs. Two aquifers underlie Valduc: one in the upper limestone (Bathonien) and the other in the lower limestone (Bajocien). The two are connected by rifts in the sector of the Combe au Tilleul. Water in both levels of limestone circulates rapidly without significant retention of substances in the water. "Thus, the site is vulnerable to possible radioactive or chemical contamination" [HC 98].

In 1997 water at most sampled points beneath Valduc showed a contamination of several 100 Bq/l. Two sampling points had activity greater than 1000 Bq/l: a sampling point near the site where tritiated waste was burned (2000 Bq/l according to one report, 4000 Bq/l according to another ) [HC 98; Bataille 97] and a point above the Combe de Noirvau. These readings represent a major decrease since 1984 to 1986, when an eastern sector of the site had readings of up to 6000 Bq/l and a southern, up to 8000 Bq/l. (The cistern near the burn site had a volumetric activity in tritium in 1997 on the order of 10,000 Bq/l, 270 times less than its 1975 high.)

Within a radius of 5 km around the site tritium levels in the ground water vary from several tens to several hundred Bq/l; within a radius of 50 km, activity is some tens of Bq/l.

At Avot (15 km to the northeast of Valduc) the Crii-Rad in 1996 found in drinking water 28 Bq/l of tritium; the CEA in 1997, 36 Bq/l. At Lemeix, also northeast of the site, the Crii-Rad in 1996 found in drinking water 63 Bq/l; the CEA in 1997, 97 Bq/l. At Lochère (1 km southwest of the center) the Crii-Rad found in a fountain 163 Bq/l; the CEA, 175 Bq/l. The Crii-Rad also detected in "other sources coming from the area of the site" up to 630 Bq/l [HC 98].

IV.C.5 VEGETATION

In 1997 the mass activity of tritium in plants was several hundred Bq/fresh kg, within Valduc; and several tens of public Bq/fresh kg outside the site [HC 98]. At the burn site, the contamination recently ranged from 190 to 2500 Bq/fresh kg [Bataille 97].

IV.C.6 FISH

In 1986 three species of fish from ponds at the site were found to have 2200 Bq/fresh kg in tritium. The CEA in 1998 did not have more recent statistics available [HC 98].