ML19225A026
| ML19225A026 | |
| Person / Time | |
|---|---|
| Site: | Atlantic Nuclear Power Plant  |
| Issue date: | 07/13/1979 |
| From: | Haga P OFFSHORE POWER SYSTEMS (SUBS. OF WESTINGHOUSE ELECTRI |
| To: | Ballard R Office of Nuclear Reactor Regulation |
| References | |
| FNP-PAL-040, FNP-PAL-40, NUDOCS 7907170180 | |
| Download: ML19225A026 (35) | |
Text
'
)
[ ll.~ FNP-PAL-040 Offshore Power Systems o w s n I mm a. u: /> : / zoo n,, anon. o , , o n.. . F M H i W1 1- ! N Fei July 13, 1979 Mr. Ronald L. Ballard, Chief Environmental Projects Branch 1 Division of Site Safety and Environmental Analysis United States fluclear Regulatory Commission Washington, DC 20555 Y,7"
Subject:
Docket No. STN 50-137; Response to NRC mm a, , , n i . ,
,o ; Re Particle Transport
Dear Mr. Ballard:
In response to your letter of June 14,19/9 to A. R. Collier, I am enclosing an assessment by Offshore Power Systems of the potential impacts that may result from the release of insoluble core debris particles following a postulated core-melt accident at an offshore Floating Nuclear Plant.
As a result of our analysis, Offshore Power Systems concludes that transpnrt of significant amounts of radioactivity as particulate matter following a postulated core-melt accident is very unlikely. If this form of transportable particles should occur, our analysis and the analogous experience (including measurements) at the Windscale reprocessing plant indicate that dose consequences are likely to be less than for the soluble radioactivity transport cases previously considered in the Offshore Power Systems and Nuclear Regulatory Commission Liquid Pathway Reports and FES-III. Our assess-ment supports tne Staff's conclusion in FES-III that considera-tion of scluble releases to evaluate consequences resulting from material released as small insoluble particles is a conservative approach.
Sincerely, f
? O ll
'd ' C - 'V 4_.
P. B. Ha,
/lel CC: V. W. Campbell A. R. Collier
^
3 5 ) 345 011 4
7 0 071i019 d
PARTICLE TRANSPORT OF CORE DEBRIS I. Introduction This ':. formation is in response to the letter dated June 14, 1979 fran Mr. Ronald L. Ballard of the NRC requesting additional informa-tion with regard to potential impacts from the release of insoluble particles produced by a core melt accident.
The Atomic and Safety Licensing Board (ASLB) for the OPS Manufac-turing License Application directed seven questions to the Applicant and NRC Staff related to the LPGS study and FES-III by their letter dated March 29, 1979. These questions were responded to by the Applicant via oral testimony at the April 4, 1979, hearing session.
At the April 4 hearing session, the IM. Staff informed the Board that the NRC responses would be at a later time and in writing due to the unavailability of Staff members at that time.
Subsequently the Staff discussed with OPS at a meeting on June 8, 1979, questions regarding (1) the formation of particulate material follcwing a core melt-through (as a result the interaction of core melt debris with basin water) , (2) dispersion of this radioactive particulate material, and (3) the magnitude, areal extent and temporal extent of dose effects resulting fran such particulate material. These questions were associated with NRC Staff preparation to address ASLB question four which is:
345 012 "4. What reasons were there for not considering interactions with sediment in off-shore cases (LPGS, p. 4-13)? Since we believe consideration should have been given, what are the effects of such interactions?"
The material discussed at the June 8,1979 meeting and the cdditional evaluations performed by OPS as a resi,lt of the June 8 discussions are documented below.
A. Liquid Pathways Reports TreatmL t of Particclate Transrurt The general approach taken in both the OPS LPGS report (Refer-ence 1) and the NRC LPGS Report (Reference 2) in estimating radiological in pact that might occur via liquid pathways as a result of postulated core melt events was to assume radioactive species remaining in the debrid would be soluble once the debris entered basin water. With this approach, holdup of radioactivity within tne basin is minimized, transport outside the breakwater is maximized as is the resulcing population dose (man-rem). In the OPS report, maximum individual doses were estimated by considering specific scenarios. OPS recognizes that other scenarios cculd be postulated to produce larger individual doses via liquid pathways; however, the cases postulated in the OPS LPGS report were considered suf ficiently conservative and unlikely by both OPS ard the Staff at the time of the re[nrts to 345 013 represent reasonable upper-bound estimates of individual dose via liquid pathways.
Dose effects resulting from transport of primary core debris particulai;e material was not treated in either the OPS LPGS report or the NRC LPGS report for the following reasons: (1) the low likelihood of an event leading to fine fragmentation of core debris, (2) uncertainty regarding the distribution of particle sizes that would be formed in the volten debris-water inter-action, and (3) the fact that a closely analogous situation was evaluated in the LPGS reports. The low likelihood of extensive fragmentation and data tegarding core debris size distributions is discusscd in Section II following. Analogous evaluations are discussed in the next paragraph.
For the analogous situation treated in the LPGS reports, either rapid release of all of the radioactivity incitrJing insoluble species fran the debris to basin water was assumed (by the h7C) or a large fractional release for all species incltriing in-solubles was assumed (by OPS) . An open breakwater with little holdup of radioactivity was also assumed. For ocean and estuarine cases, interaction between the water column and bottan sediments was considercd in such a way that there was equilibra-tion between radioactivity in the water column artl the sediment.
Thus contaminated sediment was in fact considered in these reports. The spectrum of sorbed radionuclides was like that which is present in core melt debris and this spectrum was used 345 014 in estimating dose consequences to biota, to man from ingestion of sne11riso mu to man from direct contact with ceach sedi-ments.
B. Siting Considerations NRC has required in FES-III that closed relatively impermeable breakwaters be employed for estuarine or riverine sites.
Offshore Power Systems recently proposed in the core ladle design report (Topical Report 36A59) ard in PDR Amendment 27 a more definitive wording of the site related criterion. As reworded, the criterion specifies that site .>tructutes be such as to reduce the source terms to levels equivalent to similarly sited land-based plants (source reduction of approximately a factor of 1000) . Particulate transport outside site structures is then already severely restricted for riverine and estuarine sites. With this in mind, the discussion which follows will be direc eed at ocean-sited FNP's with breakwaters which provide for relatively free interchange between the plant basin and open ocean.
II. Particulate Formation During Melt Debris - Basin Water Interaction Interaction between core raelt-debris ard basin water is discussed extensively in Sections A-2.3.4 ard A-2.3.5 of the NRC LPGS report in terms of potential for such an interaction causing a steam explosion.
NRC concludes that while an interaction involving a large fraction of
_,_ 345 015
the debris cannot be precluded, such an event is not expected. The last paragraph of the general section of A-2.3.5 states: "Considering the factors just described, it is the j udgment of the staff that energetic steam explosions and extensive fine fragmentations of molten materials wuld not be expected to occur following hull penetration by the molten material iato the salt water basin."
The staff approach to possible but unlikely outcomes of the pos'.u-lated core meltdown event is discussed in FES-III and particularly in Section 6, Resporides K-19 through K-22. In general the staff utilized the more likely scenarios as one component of the comparison of risk via Liquid Pathways of ENP's with Land-Based Plants. The other canponent was comparison of total risk derived by considering the product of probability and consequence of both the more likeiy and less likaly but possible scenarios. Consideration of dose con-sequences resulting from release and transport of fine particulate primary core debris certainly falls into the second category so that dose consequences need to be weighted by the low likalihood of extensive f ragmentat. ion. It is important to recognize that utiliza-tion of a single possible but fxprobable scenario to reach a con-clusion is not appropriate nor consistent with the approach in the NRC and OPS LPGS repor ts and in FES-III.
A. LPGS Treatment er Particulate formation and Leaching Review of the NRC LPGS reports shows that in determinirq " leach rate" fran debris material, a high rate of leaching i as used 345 016 based on equivalent 10 to 12 g diameter particles (which are very fine particles) . However, our understanding of the Staff reasoning which lead to these high leach rates is as follows:
- 1. Exporimontal data show that particles formed during Jebris basin water interaction are very porous. As a result, they exhibit a much greater effective surface area than would a hard sphere of an equivalant size. From the standpoint of hydrodynamic transport, the particle diameters are signifi-can y larger than particles sizes daich can be effectively transported out of the basin (see section I I I ', . For ex rnple, the Staff memo from R. Denise to R. Vollmer dated 3/16/77 (Reference 3) recomerds an effective particle dia:aeter of 250f4. From the standpoint of leaching, the 250/4 porous particles exhibit surface areas equivalent to that ofhardspheresinthesizerangeof10jtto20g.
- 2. Substantial uncertainty existed in the leach rate to be assigned to core debris materials. The Staff took a con-servative approach to boun3 uncertainty in the data. OPS disagreed with the Staff's use of the conservative leach rate (see Enclosure 4 to Apperdix A of Reference 2) . The high leach rates employed by the s*'f f lead to leaching c2 essentially all of the ralivactivity from the particles in a few days. With this approach, the question of particle transport is moot for particles contain little rad io-activi( and so represent very little hazard.
345 017 In summary, it is apparent from the discussion in the OPS LPGS report ard the NRC LPGS report that the physical state of the debris and the rate of lec hing following melt debris-water interaction are not well understood. However, even for the case where extensive interaction was assumed, available data indi-cates particles are relatively large frcm the standpoint of hydrodynamic transport (u250g ) . Both the staff and applicant concluded that an energetic steam explosion which might produce more extensive fragmentation and smaller particles was not expected.
B. Specific Cases Four specific cases are discussed below which cover the range frcm little or no interaction to extensive interaction and fragmentation of the core melt debris. These cases are:
- 1. Little or no fragmentation. Debris collects as a pile of large particulate rubble or a reformed slab of hot debris on the bottom of the basin. This case prcduces few par-ticles. Release of radioactivity to basin water is slow and due to leaching.
- 2. Formation of a fragmented spongy debris material with large surface area and relatively large particles. Leach rates are like those prog > sed by OPS in Reference 1. For this case, release of radioactivity from debris is relatively
_.,_ 345 018
slow. There is little if any transport of debris particles fran the basin.
- 3. Same case as 2 with high leach rates as adopted by NRC.
There is again little if any transport of debris particles from the basin. However, radioactivity is leached from the debris quite rapidly and can reach the open water body before source isolation.
- 4. Extensive fragmentation as a result of energetic interac-tion between molten-debris aM basin water to produce fina fragmentation. This case can produce particulate material with potential for transport of particles outside of basin.
C. Particle Size Distribution Resulting From Energecic Intera' _ ions (Case 4 above)
Particle size distribution data resulting from energetic inter-action between molten metals or metal oxides aM water have been reviewed by both the NRC Staff and OPS. The NRC data su:ra.ary appears in graphical form in Reference 3 (a copy of which war provided to OPS) . The OPS summary, also in a graphical form, is part of the handout for the 9/29/7) %;S Meeting and is attached as Figure 1. 'Ihe two sets of curves are similar since they are derived fran essentially the same da ta. For the discussions which follow, a conservative extrapolation of the available data to the smaller size particle range is employed to determi.ne the 345 019 fraction of the particulate material of sufficiently small size for possible transport outside the basir. for an ocean sited RJP for the unlikely case of extensive interaction between the debris arrl basin vater.
III. Particle Transport from Basin This section describes the potential mechanisms to transport par-ticles out of the basin. Tb escape the breakwater, debris particles would have to be transported from the debris on the bottom of the basin to the edge of the RIP arrl up to the surface layer of the basin water. Then they would have to be transported out of the basin before settling enough to be trapped within the basin. Although several potential mechanisms exi to transport the particles out of the basin, most can be considered neglig ible. As is shown below, only ambient currents and possibly wave motion are viable mechanisms for transporting particles out of the basin.
A. Ambient Currents Assuming the Atlantic Generating Station open breakwater configuration, a:rbient current flowing through the breakwater openings could trar, sport. particles in the surface layer out of the basin should they become entrained in the current. In its study of consequences of core-melt accidents, OPS assumed the ambient ocean current speed to be 5 cm/sec. This is the annual average current speed measured at the AGS site (Reference 4) .
345 020 The size particles which could escape the basin was estimated for for a 5 cnVsec current passing through the breakwater opening s.
To exit the basin, a particle at the edge of the alp would have to travel past the sill before it settled to the sill (6m depth) . 'Ihe minimum distance from the R4P to the center of the sill is 107 m. A particle travell.in) at 5 cm/sec would take 2140 seconds to reach the middle of the sill. Assuming a specific gravity of 4.0 (based on core debris and ladle material) for the particles, only particles less than 40 g, would be transported out of the basin. Particles greater than 40g would be trapped inside the basin. Doubling the particle drif t speed to 10 cm/sec would increase the size of particles that could escape to 567 B. Wave Motion Second order surface wave theory describes a nonperiodic drift in the direction of wave advance. Thus, waves propagating through an open breakwater would create a drif t current which could transport particles out of the basin. To estimate the potential magnitude of such a current, the AGS Open breakwater configuration was assumed. A wave of 2m in height with a period of 8 see was assumed outside the breakwater. This height is approximately the maximum significant wave hei;ht aryl the period is the most frequently observed peak spectral period for the tiew Jersey coast (Re f erence 4) . It is assumed that this wave 345 021 propagates in a direction parallel to the closure breakwater allowing it to pass directly through both breakwater openings.
Experimental data (Reference 5) indicates that such a wave travelling outside the breakwater would have its wave hei Jht attenuated by a factor of two inside the breakwater. Thus, the resultant wave propagatiry out of the breakwater would be 1 m in height. The correspanding theoretical wave drif t velocity muld be a maximum of 1.5 cWsec at the water surface and exponenti-ally decreases with depth. For this veloci- (, only particles less than 20p could escape.
For the AGS site, the larger wave heights are associated with storms an3 propagate fran the offshore direction, i .e. perpen-dicular to the closure breakwater. The experimental data indicates that such waves would be attenuated by a factor of 5 or more by the breakwater. Such waves would result in a neglig-ible drift current for transport of particles out of the breakwater.
C. Thermal Convection from Hot Debris The decay heat from the core debris on the bottom of the basin will produce a vertical convaction current which could lift debris particles up the water column to the bottom of the FNP.
An estimate of the average vertical convection velocity, V. can be calculated fran the equation (from Reference 6) 345 022
[hQ 11/3 V = 0.76 i R) where h = height at which the plume is tenninated (meters)
Q = heat source (kW)
R = radius of heat source (meters)
If the debris volume is assumed to be cone-shaped with a volume of 140 m anu naving an angle of repose of 10 , then R = 9 m.
The distance fran the debris to the bottom of the FNP, h, is assumed to be 4m. The decay heat is assumed to be 3 x 104'W K which is the magnitude expected within the first day after meltdown. The resultant average vertical convection velocity is 6.7 cm/sec.
When the particles reach the bottom of the FNP, tne thermal currents would tend to move them horizontally out toward the sides of the F11P. The horizontal velocity would be expected to be significantly less than the vertical convection velocity due to the increasing cross-sectional area in the horizontal direction ard the high degree of mixing with the ambient basin water. 'Ib estimate what size particles can migrate to the edge of the FNP, it is conservatively assumed that the particles move horizontally with a speed equal to the average vertical convec-tion velocity of 8. , cm/sec. The minimum travel distance to the edge of the EI1P for any particle is 30m. For a particle to 345 023 travel to the edge of the R& and thus have the tutential to be liftal up to the surface layer of the basin water, it must arrive at the edge of the FNP before it settles out of the horizontal current under th FNP. Assumin3 the thickness of this horizontal layer to be 2m which is half the distance from the basin floor to the bottan of the FNP, particles greater than 60 (assuming a specific gravity = 4.0) will settle before the edge of the FNP is reached. Thus, thermal convection will not make particles greater than 60g available for possible traasport outside the basin.
D. Operation of Second Unit The systems operating on the second unit (i.e. the unit not under accident con 3ition) would ten] to retard ticles from escaping the basin. Since the second unit would be shut down, the condenser circulatin] water system would not be op2 rating.
Only decay heat removal systems which have the intake and discharge structures within the basin would be in operation.
ExErrimental studies (Reference 7) have shown that the recircu-lation pattern from the decay heat removal systans create a thermocline throughout the basin. This surface thermal layer will limit the rise of the vertical thermal convection flow at the sides of the RJP from the thermal plume since the tempera-ture difference between the flow arri the ambient water will be less near the surface thermal layer. Thus, particles rising alonj the sides of the RJP would be impeded fran reaching the 346 024 e-e3
~
surface of the basin water. Since they would rise to some lower depth, this would reduce their settling depth arrl a greater number of particles wauld settle within the basin.
Thus, the decay heat removal systems of the second unit will retard rather than promote particle transport out of the casin.
E. Tidal Flushing Conservative estimates of the tidal flushing at the ICS site (period of 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />) show that the average tidal velocity is approximately 0.5 cm/sec. This estimate is based on all the tidal flow passing through the two breakwater openings. In fact, the AGS breakwater is porous and the average tidal velocity through the openings may be an order ci magnitude less. There-fore, on y particles less than a few microns could be trans-por ted out of the basin by tidal flushing. Compared to the ambient current case, tidal flushing is not a signifiant factor in particle transport.
F. Flotation of Particles By Attachment of Gas Bubbles one proposed mechanisn for transporting particles to the surface water layer of the basin where they could then possibly be transported outside the basin is attachment of non-condensible gas bilbles to the particles. Such a mechanism is not core sidered viable for the reasons discussed below.
345 025
,enksid Generally the dissolved gas content of sea water in the upper 10 to 15 meters decreases only slowly with depth and correspo:ds approximately to air saturation conditions for the surface water (for examplev 6 cc/kc for oxygen at a 20 C). In contrast, Jrs solubility increases with depth as pressure increases. For example, at a depth of 10 meters, the pressure is approximately doubled so that- twice as much gas is soluble in the water at 10 meters (30') as is soluble in the water at the basin surface.
hhil- isolated particles of debris in the size range less than 100 ja may be highly radioactive, they do not represent a large enough heat source to cause localized boiling when suspended in a large volume of water. Thus, isolated particles near the surface weald not cause boiling and stripping of non-condensible gases. It is concluded that a collection of a significant debris mass on the bottom of the basin would be necessary for boiling to occur.
If water at the basin floor is heated to near its boiling point, the quantity of gas that can be dissolved in the water falls rapidly (as the vapor pressure of the water rises) . Should boiling occur, non-condensible dissolved gases will be stripped from the heated water in the region of boiling. The steam bubbles will rapidly condense as they rise through the cooler surrounding basin water. Likewise, non-condensible gases stripped frm the water by boiling will be small and terx] to rapidly re-dissolve as they rise into the surrounding cooler 345 026 water which, at depth, has significant capability to dissclve additional gas. Thus, while boiling may occur around collected debris at the basin floor, neither the steam bubbles nor stripped non-condensible gases would be expected to rise more than a few meters before they diaappear.
Also considered was the passibility of gas generation near the particles due to radiolytic decomposition of the water from radiation emitted by the particles. Gases so formed (11 2 and 0) 2 at depth can of course be rapidly dissolved by surrounding water. Metallurgical experience in flotation processes indicate that oxide materic's tend to be hydrophilic (air bubbles do not attach) unless they are coated 's i oils or organic materials which make them hydrophobic (silicates beitig an exception) . The high temperature of the oxide debris material imediately prior to its contacting sea water would destroy any oils or organic material present in the debris material. It is concluded that the debris material would discourage gas bubble attachment to debris particles.
G. Postulated Steam Explosions A steam explosion under the RD from debris-water interaction would create a bubble which would not reach the edge of the FNP (Reference 1). IJpon collapse of the bubble, the particles would return to their approximate initial location. Thus, steam 34502"(
explosions are not a significant mechanism to transport parti-cles out of the basin.
IV. Particle Dispersion Outside the Basin This section describes the dispersion of particles of core decris that escap the basin. Basca on the discussion of Section III above, any particles that can escane the basin will be less than 40g . Based on the prticle size distribution discussed in Section i'l above, particles less than 40g would account for no more than 10% of the core mass and probably less than 1% of the core mass.
A. Initial deposition F
Table 1 decribes some of the physical characteri.; tics of log and 40 y particles. For each size, it is assumed that the particles make up 10% of the ccre mass ar 6 x 10 6gm and 10% of 8
the core activity or 10 Ci. The Stokes settling velocity is 13m/ day for the 10pparticles and 208m/ day for the 40p particles.
The Stokes settling velocity is a discrete settling velocity, and if the particles settled u , a cloud rather than es discrete particles, they would have a higher effective settling velocity.
Assuming that the particles do get out of the breakwater, they will be transported in the near field by currents until they settle on the bottom, and then will be transported along the coast ani off the shelf as sediment bed load. Based on the 345 025 particle settling velocities in Table land a water depth of 10m, the currcnt transport will be ccepleted in about a day. The shelf sediment transport will take place much slower. Since the shelf sediment transport is a dispersive process, the maximum bottam concentrations (no. particles /m or Ci/m ) will occur after initial deposition and then decline due to sediment transport.
Assuming a drift current of Sem/sec (4. 32 km/ day) , the 10g particles would be deposited on the ocean sediment over a distance of less than 5 km from the breakwat.er. The 40g particles would settle within 200 m of the breakwater.
B. Long Term Sediment Transport Particles of sizes from 10/4 to 40ft may be transported from the basin and initially settle to the botton sediments in the region near the breakwater. The particles may then be resuspended duriry storm events arr3 transported and dispersed over a greater area to lower particle densities (no. particles /m2) . In general, the coastal transport of particles is a dispersive process with the particles having lower and lower densities on each succes-sive stage of transport. Their maximum benthic density would be after the first stage of settling and before resuspension by a subsequent storm. Thus, biological uptake and doses will be at a maximum af ter initial settling.
345029
/
Since the 10 4 particles are similar in size to fine silt found on the continental shelf ard the 40g particles are of the size of fine shelf sands, it is expected that the particles would be transported in a manner similar to these two components of shelf sediment. The fine silt and fine shelt sands are transported by different mechanisms that represent a spectrum of processes which transport them along the coast and off the shelf.
The larger sized particles would participate in the bed load transport of fine sands. Studies of transport of shelf sands have been corriucted in the RIST (Radio Isotope Sard Tracer) experiments (Reference 8) . Sands which are resuspended during storm events are transported and dispersed by the drif t cur-rents. Because of their high settling velocities (approximately 200Wday) they are rapidly redeposited on the bottom. Resus-pension typically takes place every two to three weeks and has a duration of about one day (Reference 9) . Some of the material can be reworked to depths of 10cm to 20cm by benthic organisms and may not be transported by each storm event.
Portions of the fine shelf sands are transported along the beaches. Coastal sediment transport rates as high as 3
100,000m / year along the Delmarva Coast have been estimated 5
(Reference 10). This transport rate represents about 10 Wes the maximum number of 40g particles that might be released and so is representative of the order of minimum dilution of the particles if they initially participated in coastwise transport.
345 030 The exchange of coastal sanls with sands in estuaries was investigated by examining the chemical composition of silt and sands in the mouths of estuaries and was summarized in Reference
- 11. There is little evidence for net transport of sard into the mouths of estuaries. As much of the fine sand must be ejected fran an estuary as enters an estuary, usually during storm events. Fine sands that do enter an estuary, are not transported very far up the estuary.
The small particles will be transported with the fine silt. The cross-shelf transport of fine silt have been studied in Refer-ence 12. The main source of fine silt is the residual of river silt loads initially deposited in the estuaries along with residual organic matter. Silt " plumes" are found to extend out of the mouths of most estuaries and are re-entrained on succes-sive tides. The general tendency is, however, for the fine silt material to have a net transport outward from the estuary and across the shelf rather than deposition of costal silt in the estuary.
V. Radiological Effects A. Dose to Biota The analysis of dose to crustacea in the OPS IAGS report (Reference 1) showed the area of mor tal t . - the largest release case was 13 km in length and about 2 km wide. As 34503$
discussed in Section IV above, particles of 10 g or larger would be deposited within 5 km of the breakwater assuming a drift current of 5 cn/sec. Even selecting the largest monthly resultant surface current measured at the AGS site during 1973-1974, which is 14 an/sec., the particles 10g and larger would be deposited within 13 km of the breakwater. Assuming the entire area of particle deposition was lethal to crustacea, this area of mortality is about the same as that calculated in Reference 1 for the most severe soluble radioactive release.
B. Dose to Man The dose to man from seafood ingestion assuming totally soluble radionuclides was calculated in Reference 1 using the bioaccumu-lation factor method to estimate radionuclide concentrations in edible portions of fish and invertebrates. Since this method is not applicable for insoluble radionuclides in particulate form, the dose to man is evaluated by calculating the dose per particle and then considering the Intential for ingesting the particles from edible portions of fish an3 crustacea.
Some of the radionuclides in the particles are soluble in sea water arr3 will be rapidly leached leaving only the less soluble nuclides. The nuclides remaining in the particles can be determined by the leach rate factor, F (Reference 1, Appendix J). Based on consideration of total activity released, solu-bility and the radiological characteristics of specific 345 032 nuclides, the nuclides listed in Table 2 were determined to be potentially impor tant. The resultant doses fran ingesting particles of 10p and 40/4 are also given in Table 2.
For an ingestion dose to man, particles containing radioactivity must be retained in the cdible portions of seafood. In the case of both marine fish and invertebrates, there would be little transfer of insoluble materials from the gut to edible tissue.
Radioactive particulates cannot pass directly through membranes and be absorbed in tissue in fish (Reference 13). Direct ingestion of sea water is a more effective mechanism for accumulation of radionuclides in marine fish and invertebrates than ingested fcN (Reference 14 and 15) .
Since there is little transfer of insoluble radionuclides to edible portions of fish, the fish ingestion pathway to man is r.o t considered to be a signif' : ant do; ,uthway for particulate matter. In addition, the fish irgestion doses calculated in Reference 1 assuming soluble releases of radioactivity would be expected to be greater than doses resulting from ingested perticulate matter.
Some marit:e biota cuch as oysters, quahogs, mussels and soft shell clans which are eaten whole would be a potential dose pathway to man since any insoluble particles of rclioactivity contained in the organism would be ingested. These organisns feed on suspended or dep] sit material and have the capability to 345 033 reject particles on the basis of size, density or ind igest-ibility (References 16 and 17) . If the particles are ingested by the organism, removal of the organism from the contaminated envirorrnent would result in the depuration of the organism in one to two days (Reference 18). This removal process would afford adequate interdiction to minimize the dose to man.
As discusseu in Section IV above, the maximum concentration of activity will occur after initial deposition and result in an area of mortality to biota. Assuning source interdiction to reduce the nu:nber of particles by a factor of 100 and a dilution 5
factor of 10 as discussed in Section IV above, the maximum particle density would be expected to be no more than 39 particles /m . Assuming a person ate a bivalve containing all the ra3ioactive particles in a square meter with no depuration, the GIT dose muld be 14.4 mrem. This is three orders of magnitude less than individual doses calculated in Reference 1 for soluble releases.
C. Effects of Particles on Population Dose Analysis of the effects of sedimentation on population doses for estuarine sites showed that for periods less than 100 days, the effect of sedimentation was to reduce the ppulation dose (Reference 2) . For times greater than 1000 days, the population doses with and without consideration of sedimentation were approximately equal. Thus, the population dose, when integrated 345 03&
over relatively long times, is almost entirely depervLnt on tbc activity released. By analogy, the effect of assuming a partic-ulate form of radioactivity rather than soluble radioactivity is to reduce the population doses for shorter times.
D. Wiidscale Experience The ex pe r ience at Windscale, England is analogous to the postulated e enario discussed here. The fuel reprocessity plant at Windscale, England has discharged plutonium nuclides in its 4
effluent to the Irish Sea for many years. Approximately 10 Ci of Pu 239 were discharged between 1967 and 1974. It is estimated that 96% of the plutonium di charged to the water is deposited to the sediments in the immediate v Oinity of the outfall.
(Reference 19)
Measurements of the plutonium concentration in the sediment irdicate that the concentrations are about the same order of magnitude within 9km of the outfall.
Fran these measurements, it was concluded therefore that the fraction of plutonium which leaves the water does so because it joina the particulate phase, which in turn is being dispersed with the general sediment load. In this respect, other fissica products such as Rul06 arrl Cel44 are found to behave in a
, similar fashion to plutonium. Following their discharge to sea, a
these nuclides, like plutonium, are lost rapidly fran the water
{
4 345 035
. i phase, but their concentration in sediment relative to the prevailing water concentrations shows very little enhancanent in the vicinity of the outfall compared with the concentration at distance.
The concentrations of plutonium in the important food materials in the Irish Sea have been evaluated m3 the safety of past discharges in a public hcalth context has been clearly demon-strated. No evidence of build-up cf plutonium in any material has been found and concentrations in sea water, sediment and biological materials have been related by a constant factor to the rate of discharge of the nuclide.
In the postulated core-melt accident, em with the highly con-servative estimate of the number of particles that escape the basin, interdiction using dredging could reduce the remaining 4
activity in the sediments to levels of the order of 10 Ci.
Based on the experience at Windscale, no significant public health effects would occur due to assumed particle transport of radioactivity out of the basin.
VI. Interdiction Interdiction to mitigate the consequences of debris particle deposi-tion outside tre basin would be feasible. Monitoring would readily identify the areas of high concentrations of radioactivity.
345 036 As discussed in section IV above, the 10 /4 particles would be initiali deposited relatively near the breakwater depending on the ambient current. Based on coastal hydrological current data, the largest area of deposition for 10/4 particles would Le expected to be about 25 km2. For 40 f4 particles, the area would be an order of magnitude less. Even an area as large as 25 square kilometers could be dredgcd to eliminate this contamination. Since radioactivity in the sediment would be expected to be confined to the top 10 an of the seabed during the first year, the total volume of sediment to be dredged over this contcminated area would be 4 million cubic yards.
Dredging millions of cubic yards in ocean depths of 40 to 50 feet has been 'ccanplished. For example, the beach erosion control project in Duval County, Florida involvcd dredging 3.3 million cubic yards of sand in 40 to 50 foot depths off the Atlantiu coast and depositing the sand along 10 miles of shoreline (Reference 20) . The project cost was $]0,578,000. Thus the cost of dredging a contaminated area of 25 square kilometers would be expected to be about 13 million dollars.
This cost can be compared to the 22 million dollars estimated for source interdiction for a land based plant (Reference 21) .
VII. Summary A conservative evaluation of the mechanisms available to transport contaminatcd particles out of basin shows that it is unlikely that particles greater than 40 g could escape the basin. An upper bound of 60 p for escape is estimated. Both NRC ard OPS concluded in the 345 037 LPGS studics that an energetic interaction between molten core debris and bast' water to ,1 ue . fine particulate material is unlikely.
Available dat a F.how that should such an interaction occur, about 1%
of the <'ebris .it form small rarticles, with 10% as an upper bourd.
For evaluations corducted here, it is assumed that 10% of the debris is fragmented to particles small enough to be transported by currents. Particles in range of 40 g to 60 p. which do escape the basin would settle out of the water column within 200 meters of the breakwater.
Smaller particles, such as 10 p, can be expected to be transported out of the basin ard would be initially deposited on the sediment within about 5 to 10 km from the breakwater depending on the ocean currents. Source interdiction usiryg dredging can almost completely eliminate potential dose consequences to man from particulate matter.
The area of initial deposition of the 10 g particles that escape the basin could produce an area of mortality to marine biota as large as 25 km 2 which is the same as the area calculated for the most severe soluble release case in Reference 1. However, after dredging of this contaminated area, any contaminated particles remainirg would be diluted by sediment transport to levels which would not result in significant doses to biota or man.
The of initial deposition of larger particles of 40 t fwould produce a much smaller area of mortality. In add ition , since these larger particles are slower to disperse, more ef fective dredging would be 345 033 possible. In Mdition, the sorting mechanism which occurs within organians that are part of the food chain to man tend to eliminate these larger particles as a significant source of dose to man.
The basis for canparison of FNP and land-based plants in the 1H0 LPGS report was population dose. Since the population dose depends on the amount of radioactivity relcased to the environment, population dose to man is not affected by release of part of the activity as particulate matter unless there are interdiction measures. With interdiction to clean up particulate matter, the ppulation dose would be reduced. Effective source interdiction is expected to be easier to accomplish if radioactivity is released as particulate matter as comparcd with release in soluble form.
A review of the analogous experience at the Windscale reprocessing plant supports the conclusion that no significant dose to man would occur from the particle releases assumed here and with feasible interdiction.
OPS concludes that transport of significant amounts of radioactivity as particulate matter following a postulated core-melt accident is very unlikely. Further, if this form of transportable particles should occur, dose consequences are likely to be less than for the soluble radioactivity transport cases previously considered in the OPS ad NFC LPGS reports ad FES-III.
345 036
- OPS agrees with the Staff conclusion (see Section 5.3.4 of FES-III) that consideration of soluble releases to evaluate consequences is a conservative approach.
345 040 REFERENCES
- 1) Of fshore Power Systems, " OPS Liquid Pathway Generic Study," Topical Report No. 22A60, June 1977.
- 2) U.S. Nuclear Regulatory Commission, " Liquid Pathway Generic Study:
Impacts of Accidental Ralioactive Releases to the flydrosphere from Floating and Land-Based Nuclear Power Plants," NUREG-0440, February 1978.
- 3) NRC Memorandum from R.Denis to R.Vollmer, Assistant Director for Site Analysis, " Floating Nuclear Plant-Core Melt Release Evaluation",
March 16, 1977.
- 4) U. S. Nuclear Regulatory Commission, Draft Environmenta' Statement for Atlantic Generating Station Charts 1 and 2, Docket Nos. SW 50-477 and STN 50-478, NUREG-0058, April,1976.
- 5) Public Service Electric & Gas Co. of New Jersey, "Model Tests to Evaluate the Atlantic Generating Station Platform Response to Periodic Waves of Varying fleight, Direction ard at Various Water Levels", Preliminary Safety Analysis Report for Atlantic Generating "tation Appendix 30, February 1976.
6 Personal communication from G. R. Hadley of Satr3ia Laboratories to Gordon Chipnan of NRC (supplied to OPS by NRC) .
- 7) G. Jerka, D. Wood and D. Harleman, " Theoretical and Experimental Investigation of Emergency Heat Releases from Floating Nuclear Fawer Plants", Report No. 206, Massachusetts Institute of Technology, November. 1975.
- 8) Duane, D. B., " Synoptic Observations of Sand Movement", Proc. 12th Coastal Engineering Conf. , Washington, D.C. (pgs. 799-813), 1970.
- 9) Swift, D. J. P. and J. C. Ludwick, " Substrate Response to Hydraulic Process: Grain-Size Distributions ard Bedforms", in Marine Sediment Transport and Environmental Mana'ement. J. Wiley and Sons, N. Y.,
1976.
- 10) Moody, D. W., " Coastal Morphology and Processes in Relation to the Develognent of Sutmarine Sand Ridges of f Bethany Beach, Delaware."
Ph.D. Thesis, John Hopkins University, 1964.
- 11) Meade, R. H. , " Landward Transport of Bottom Sediments in Estuaries of the Atlantic Coastal Plain", J. Sediment. Petrol., 39:222-234, 1969.
- 12) Schubel, J. R. arrl A. Okubo "Comnents on the Dispersal of Suspendal Sediment Across the Continental Shelf" in Shelf Sediment Transport:
Process and Pattern.
345 040
.s
- 13) Hoss, D. E. and J. P. Baptist 1971. Accumulation of Soluble and Particulate Radionuclides by Estuarine Fish. In Proceedings Third National Symposian on Radioecology. USAEC-CONF-710501. pp. 776-782.
- 14) Penetreath, H. J., D. S. Woodhead and D. F. Jefferies 1;71.
Ra3ioecology of the Place (Pleuronectes Platessa L.) in the Northeast Irish Sea. In Proceedings Third National Symposium on Radioecology.
USAEC-COW -710501, m. 731-737.
- 15) Guary, J. C. and A. Fraizier 1977. Influence of Tropic Level and Calcification on the Uptake of Plutonium Observed. In Situ, in Marine Organisms. Health Physics 32. pp. 21-28.
- 16) Raid, R. G. B. 1971. Criteria fur Categorizing Feeding Types in Bivalves. The Veliger. 13.
3 pp, 353-359.
- 17) Prosser, C. L. (ed), 1973. Canparative Animal Physiology. Third Edition. W. B. Saunders Co. , Philadelphia.
- 18) Foster-Smith, R. L. 1975. The Effect of Concentration of Suspension and Inert Material on the Assimilation of Algae by Three Bivalves.
J. Mar. Biol. Ass. U. K. 55. pp. 411-418
- 19) He ther ing ton, J. A., Jefferies, D.F. and Lovett, M. B., "Some Investigations into the Behavior of Plutonium in the Marine Environment", Impacts of Nuclear Releases into the Aquatic Env ironment, International Atanic Energy Agency, Vienna, Austria, 1975.
- 20) U. S. Corps of Engineers, Final Environmental StatemerM, Beach Erosion Control Project, Duval County, Florida, August,1974.
- 21) U. S. Nuclear Regulatory Commission, Final Environmental Statement, Part III, NUREG-0502, December ,1978.
345 042 i .
1 h
~
10,000 _
1
- g I /
4,000 -+ - - - -- - -
y -
g-~
'l /
If
/1 2,000 - T-
-~
~
~,/9;d- - ~~ - -
-I '
,/) <
l +-
M 4 .f -- -
z 1,000 3i ' "C o - / 3h
- f'2, j l J,/
m
- l
/ f/
d
/$ l
~
400 ilr
- m-
,/ ,
- l .
a I ' '-
y 200 i i/ ,
l l / liI !- I l l d
I
/ I b l I I 2- 100 l /
1-ARAKERI (UCLA) 50 2-GIDBY
~ 3-SA:lDI A 30 PARTICLE SIZE DISTRIBUTIO::S OF GL.iS3/H2 0,
'!D CORIU:I/H G g EXPERI: TESTS
~
UOg /H gO I - ** i i i i i uu w >> un.,
i uu ' ' ' '- *- 'i et 10 i
.1 1 10 30 93 30 95 90 90.0 09. 0C
.01 17EIGHT PER CENT FI:iER Fi ! ;ure ]
32 345 04B
r Table 1 Particle Characteristics for 107,of Total Core Total mass (10% of Core) Gx10 gmE Total activity (105 of Core) 1x10'Ci Specific gravity 4.0 gm/cm 3 Particle Size, p 10 40 Particle Weight, mg 2x10 6 1.3x10 "
Number of Particles 3x10 15 5x10 13 Settling Velocity (cm/sec) 0.015 0.24 Settling Velocity (m/ day) 13 208 33 345 044-
r l
a n
r e
t
) n I
E L r s C )
- o s I M ) f e T E ) ) ) ) ) 2 ) r R R 1 0 0 1 1 - 0 e P A M m ( ( ( ( ( ( ( s P ( c o n 9 1 0 8 6 1 7 D o R E 0 . . .
m E L 4 1 7 6 1 3 3 6 e a P C 8 l g I
b r M T i e E R s P R A s M P i ,
(
R m 2 S E r
E P e n T
P o i
A T n t L I o a U G
- c C ) ) ) ) ) ) I i I O 1 1 1 1 1 4 ) I l T T - - - - - - 0 b R m ( ( ( ( ( ( (
e u A E u 4 e P P S 1 2 0 0 1 2 4 t O 0 . .
t ,
S D 1 3 1 1 3 6 5 1 i n I
R m o B m i E
o t D
C c e
P t E R o R C r O ) ) ) ) ) ) ) I P C i 1 1 1 1 1 4 C - - - - - - f l N p
- m. ( ( ( ( ( (
o a I ( .
4 4 8 0 2 0 c
t i S E 0 r g E L 4 2 3 2 2 5 D C 1
o p
o l
I I L T e o R i C R d U A e a N P h R O t I R n D E f o A P o R n Y ) ) ) ) ) )
y o M T 3 3 3 3 3 6 g i O I - - - - - - o s R V m ( ( ( ( ( ( l s F I u o i T 0 7 7 7 7 3 d m E C 0 . . . .
o m S A 1 4 5 4 1 3 8 h o O t C D e T m l I t a G c o n a t o i
2 r g t t n E a L
i n B
l a d r n r e A o t T i c n t
s c I e a t d .
E n e n D i t o I -
L 3 o a i 6 4 8 l t C 5 0 0 4 3 r u a U 1 9 1 1 1 2 t c i N 9 - - - - - s d 1
u4LJ D4N l
O - b u u o u L a a a I Y N R R C P A G C R C D T A O R T * *