0160 4120 (84) 90054 0
0160 4120 (84) 90054 0
0160 4120 (84) 90054 0
00
Printed in the USA. All rights reserved. Copyright © 1985 Pergamon Press Ltd.
A comparative overiew is provided of some major aspects concerning assessment and management of
chemical and nuclear risks arising as a consequence of accidents. Statutory procedures for constructing
and running nuclear and chemical plants in Italy are discussed in detail. Special attention is given to the
major changes that are likely to occur after the adoption of the EEC Council Directive 82/501, designed
to prevent major accidents which might result from certain industrial activities and to limit their conse-
quences for humans and their environment. Present status and future trends of accident analysis and risk
assessment are also been dealt with, and special emphasis is placed on aspects common to both nuclear
and chemical plants. Lastly, managerial aspects of contingency planning for, and response to, emergen-
cies and accidents involving toxic chemical and/or ionizing radiations are examined with the aim of iden-
tifying more critical steps.
in consultation with the Technical Commission, issues by the technical advice of local competent health units
its authorization for plant trials to begin. However, (established under the National Health Service Act in
these trials can only be authorized once an off-site 1978).
Emergency Plan has been formulated by the local Pre- If necessary, local authorities can obtain further
fetto in collaboration with the other competent bodies technical and scientific support from two central Institu-
on behalf of the Ministero dell'Interno (Home Office). tions, the Istituto Superiore di Sanit~t and the Istituto
Once plant trials have been completed, ENEA, hav- Superiore per la Prevenzione e la Sicurezza del Lavoro
ing consulted the abovementioned Technical Commis- (National Institute for Prevention and Safety in Work),
sion, then forwards to the Ministero delrlndustria its through appropriate channels. All chemical plants must
opinion on the issuing of the licence for plant operations, comply with several regulations concerning occupational
along with the technical prerequisites to be fulfilled. For and environmental hygiene, as well as accident and fire
nuclear power plants, apart from this procedure, Law prevention. In particular, insalubrous industries must
No. 393 (from 1975) and Law No. 8 (from 1983), which comply with articles 216 and 217 of the R.D. 27 July
stipulate site selection criteria, also apply. According to 1934 No. 1265; these articles state that plants that pro-
these provisions, it is up to the CIPE (Interministerial duce insalubrous emissions or that could be otherwise
Committee for Economic Planning), in agreement with dangerous to health of the population are to be classified
the Commissione Interregionale (Interegional Advisory according to two lists or classes. Class I consists of
Commission), to decide which regions may host power plants to be isolated from inhabited areas; Class II con-
plant sites. It is therefore the regional authorities' sists of plants that require special preventive devices and
responsibility to indicate at least two areas for each measures. For public health reasons, local authorities
plant within 150 days of the CIPE decision. This must can refuse licencing of a plant or impose specific norms;
be done in agreement with the municipal authorities they also must ensure that the plant complies with ex-
concerned, and with the technical assistance provided isting regulations.
by ENEA, after having acquired the opinion of the Recently, an EEC Council Directive has been approved
Health Ministry and having consulted the National (EEC, 1982) to prevent major accidents which might
Electricity Board. result from certain industrial activities and to limit their
Sites are selected, out of those proposed, on the basis consequences for people and the environment. When
of a "qualifying process" which ascertains the technical implemented in Member Countries, this Directive will
suitability of each site. These enquiries form part of a represent a major advance in preventing health risks
technical investigation and are submitted to ENEA and arising from chemical accidents (Sampaolo, 1980). Ac-
competent Ministries, as well as the Technical Commis- cording to the "frame part" of the Directive (article 3
sion. Lastly, with the investigation completed, the and 4), Member States shall adopt the appropriate pro-
regional authorities select the site. If the procedure to vision to ensure that the manufacturer is obliged to take
select nuclear power plant sites has not been completed all the measures necessary to prevent major accidents
within the time stated by Law No. 393, Law No. 8 stipu- and to limit their consequences for people and the en-
lates that selection of possible sites shall then be per- vironment. Moreover, Member States shall take the
formed by CIPE acting on proposals from the Ministero measures necessary to ensure that all manufacturers are
dell'Industria. required to prove to the competent authority, at any
In this licencing scheme, three major points must be time, that they have identified existing major-accidents
stressed: hazards, adopted the appropriate safety measures, and
1. The technical responsibility of the licencing pro- provided the persons working on the site with informa-
cedure belongs entirely to national authorities, whereas tion, training, and equipment in order to ensure their
local and regional authorities play a key role in site safety.
selection only. The "specific part" of the Directive deals with norms
2. The law itself specifies, in a rather detailed way, the concerning any operation carried out in an industrial in-
main contents of the plant safety report. stallation referred to in the Annex I of the Directive, if
3. Apart from plant evaluation, a site licencing pro- one or more of the dangerous substances listed in the
cedure is also foreseen. Site selection must be performed Annex III are involved, or if it is recognized that they
from a list of potential sites drawn up under the respon- may be involved in the quantities and modalities laid
sibility of the ENEA. Obviously, except for possible down in the said Annex. The "specific part" applies also
minor adjustments, the site selection list for nuclear to any operation carried out in an industrial installation
power plants, established in Italy since 1979, could also if one or more dangerous substances listed in the Annex
be suitable for other high-risk nuclear or non-nuclear in- II are stored in the quantities laid down in the second
dustrial installations. column of the same Annex. In all the abovementioned
cases, Member States shall introduce the necessary mea-
Chemical Plants sure to require the manufacturer to notify the compe-
In Italy, the responsibility of licencing chemical tent authorities with (1) information relating to relevant
plants belongs to local authorities, who are supported substances; (2) information relating to the installations;
Chemical and nuclear risk management 477
and (3) information relating to possible major-accident before the industrial activity starts. The purpose of the
situations. In the case of new installations, the notifica- notification is to provide information relating to sub-
tion must reach the competent authorities a reasonable stances, installations and possible major-accident situa-
length of time before the industrial activity starts; the tions.
Directive shall also apply to existing industrial activities. 4. According to the abovementioned Directive, the
It is worth noting that this mechanism should allow a competent Authorities will: (a) examine the information
rapid effective identification o f existing high-risk ac- received with notification; (b) ensure that an emergency
tivities, which is strongly needed at the moment. plan is drawn up for action outside the establishment;
The Directive also establishes an information (c) organize inspections and other measures as necessary;
mechanism in cases of accidents. As soon as a m a j o r ac- (d) collect, when possible, the information necessary for
cident occurs, the manufacturer shall be required to in- a full analysis of the m a j o r accident and possibly make
form the competent authorities about the causes o f the recommendations.
accident and the anticipated steps to alleviate its medium- 5. Lastly, an information mechanism on chemical ac-
and long-term effects and to prevent any recurrence of cidents occurred will be in operation.
such an accident.
F r o m their side, Member States shall inform the In conclusion, the EEC Directive, once adopted by
Commission and the Commission shall set up and keep Member States, could work for some types of chemical
at the disposal of the Member States a register contain- plants as a licencing scheme.
ing a s u m m a r y o f the m a j o r accidents. This summary
shall include an analysis of their causes, experience
Present Status and Future Trends of Accident
gained, and measures taken, to enable the Member
Analysis and Risk Assessment Applied to Nuclear
States to use this information for prevention purposes.
and Chemical Plants
It is a responsibility o f Member States (article 7) set-
ting up or appointing the competent national authorities
Nuclear Plants
who are responsible for (1) receiving the notification
As already mentioned, authorization to build a
and any information; (2) examining the information
nuclear power plant is tied to approval of a safety report
provided; (3) ensuring that an emergency plan is drawn
analyzing possible plant malfunctions and consequences
up for action outside the establishment in respect to
in terms o f releases to the environment. A m o n g the var-
those industrial activities for which notification has
ious potential accidents, m a j o r attention used to be
been given; and, if necessary, (4) ascertaining that the
given to the analysis o f one or more accidents of con-
manufacturer takes the most appropriate measures, in
siderable seriousness from which, on one hand, engineers
connection with the various operations involved in the
could gain sufficient data to dimension plan safety
industrial activity for which notification has been given,
features and, on the other hand, control authorities
to prevent m a j o r accidents and to provide the means for
could assess plant compatibility with the selected site.
limiting the consequences thereof.
However, over recent years, there has been a gradual
The competent authorities shall organize inspections
shift in Italian policy, going from a deterministic acci-
or other measures of control proper to the type of activity
dent analysis to a probabilistic type approach. The
concerned, in accordance with national regulations.
adoption of the latter has enabled the introduction of
In conclusion, although the present licencing scheme
so-called design objectives, whereby estimated annual
of chemical plants in Italy will change to some extent
probability of anomalous events can be correlated with
when the EEC Directive on m a j o r chemical accidents
the m a x i m u m permissible radiation doses for the popu-
will be adopted, the situation m a y be summarized as
lation at large.
follows:
For example, in the case of the plant being built at
Montalto di Castro, design objectives have been fixed as
1. The licencing procedure is mainly under the
shown in Table 1. A close look at the table shows that it
responsibility o f local authorities and their technical
agencies. Advice is available through appropriate chan-
nels f r o m the central institutions of the National Health Table 1. Design objectives for the Alto Lazio (Montalto di Castro)
Service. nuclear power plant.
2. Once the EEC Directive on m a j o r chemical ac- Event Probability p Total Body Dose
cidents is adopted, the manufacturer will still have the (per yr) Limits to the Public
responsibility for identifying major-accident hazards
Routine releases - 10-~ m Sv
and for adopting the appropriate safety measures and Anomalous 5 x 10-2 _< p < 1 10-~ m Sv
providing the persons working on the site with informa- transient and 10-~ -< p _< 5 x l0-2 5 m Sv
tion, training, and equipment in order to ensure their accidents
safety. PosA a <- p <- 10-a 10-1 Sv
3. Moreover, in cases of high-risk activities, a aThe most severe Design Basic Accident can be thought as a loss of
notification to competent authorities must be presented coolant accident with adequate emergency core cooling.
478 G . C . Venuti et al.
also refers to normal conditions but does not contain Table 2. Different scenarios for light-water reactor accidents.
limits for accidents any more serious than the designed
Fraction of core
basis ones. However, the prevailing tendency of the con- inventory released
trol authorities in drawing up the general design criteria Probability
for the package of six 1000-MWe units, foreseen by the (reactor x yr)-' Noble gases Iodine
Italian National Energy Program for 1990, is that of Large scale accident:
specifying a probability ceiling also for the events that PWR-2 release category
entail serious damage to the fuel elements. In fact, the o f W A S H - 1400 Report
overall annual probability of going over fuel cooling (NRC, 1975) 10-s-10-6 0.9 0.7
Large scale accident:
capacity must not go beyond a given value in the range
revised estimate 10-~ 0.6 10-~
of 10-5-10-6 per reactor. Design basis accident 4 x 10-4 3 x 10-~ 10-7
Requiring plant engineers to adopt a probability ap-
proach to safety criteria is, however, not limited to en-
suring that objectives such as those listed in Table 1 are
complied with, but also entails systematic checks on reassessment of the evolution of the most serious acci-
plant behavior during accident situations, together with dent sequences and of a detailed analysis of experimen-
an assessment of safety system performances for events tal data concerning chemicophysical control mechanisms
with a degraded core. In particular, it is required that of release of relevant radioisotopes. However, a large
plant engineers, when calculating event probabilities, amount of work is still in progress in order to clarify all
also take operator reliability into account. Various the items of this reassessment. The results of such a re-
countries over the past few years have been extending evaluation has been sketched in Table 2, where both the
the accident scenario for safety assessment so as to have major and revised accident scenarios are presented to-
overall nuclear power plant risk assessment. In connec- gether with design-basis accident parameters.
tion with this fact it may be worth remembering that the The reduced fraction of the iodine isotope inventory
criteria recently adopted in the United States (NRC, released is essentially due to the iodine retention mech-
1983) refer both to individual and overall risk ceilings as anism discussed by Becher et al. (1981), Gjorup et al.
well as to plant performance guidelines for large-scale (1982), and Morewitz (1981).
core melts. In these guidelines, the probability limit for In Fig. 1 the individual doses from the two major ac-
large-scale core melts is 10-'/yr per reactor. cidents are compared; the figure shows the 30-yr com-
Adoption of design objectives entails limits that are mitted effective dose equivalent of the two different
not only restricted to plant performance but also ac- releases. At distances of 2-3 km or more, the revised
count for population settlement patterns around plant doses are about a factor of 100 lower than those from the
sites. In particular, the Italian procedure for licencing hypothetical large accident (WASH-1400 estimate;
makes reference to two site-quality indexes. The first is NRC, 1975). For the acute dose the differences are
calculated by integrating the population within 20 km smaller; at distances of 2-3 km or more the lower doses
around the site, and is weighted with a suitable atmo- are about a factor 10 less than those from the worst
spheric dilution factor. The second index is calculated scenario. This is due to the dominant position of the
integrating, within the 22.5 ° most inhabited sector, resi- dose from noble gases in the revised case with respect to
dent population weighted with a dilution factor that the reduced iodine releases.
goes as r -3'2, where r is distance from the plant. Integra- Analysis of releases associated with the design-basis
tion is performed up to 50 km from the plant, thereby accidents suggests some minor health implications ex-
providing some indication of the overall exposure of the tending not more than 2-3 km from the plant. It should
population within such distances. The two indexes thus seem advisable to carry on with the present policy
calculated must be compared with two reference values adopted by the Italian authorities, who require that the
taken as limits for assessing site suitability. Procedures interested areas be limited to farming activities with
adopted to fix these values are described in CNEN- small population settlements. If acute consequences on
DISP (1977). health of catastrophic accidents are also to be minimized,
It is worth noting that the WASH-1400 report ( N R C , a larger area should be kept under control (Technical
1975) assigns a probability per reactor-yr from 10-4 to Commission for Nuclear Safety and Health Protection,
10-s to those event sequences that could potentially lead 1978). Bearing in mind average weather conditions, the
to core meltdowns. Among these sequences are those relevant distance from the plant should be of the order
entailing a substantial loss of containment, which could of 10 km.
implicate an external release of considerable quantities At any rate, if the conclusions of the reassessment of
of radionuclides. Probability of such events has been the most serious accidents would be confirmed, this
estimated of approximately at 10-s-10 -6 per reactor-yr. would imply going from a catastrophic accident scenario
The reliability of these estimates and those of respective to one with 10-100 immediate casualties under adverse
external releases have been re-examined on the basis of a weather conditions. The area implicated in determining
Chemical and nuclear risk management 479
10-11
three categories together with the relative consequences
in terms of fatalities. Only the first category, which ap-
pears to be more directly comparable with the radio-
active release phenomenology, will be discussed here.
The most part of the relevant accidents turns out to be
associated with chlorine releases. Nineteen accidents are
10-2- recorded in the period 1917-1979. The number of fatal-
ities in each event has decreased starting from the 1940s.
The historical risk of chlorine releases can be calculated
in terms of the amount of chlorine which gives rise to
10-3t,,,,I I L , I,,,,I, ,
one fatality; the mean mortality index, so defined, is
0.5 1 5 10 50 equal to 0.3 fatalities per ton of chlorine released.
Although this figure allows for an approximate chlorine
distance downwind / km source risk evaluation, it should be noted that the effec-
tive value of the mortality index crucially depends on (a)
Reactor power: 2000MW t the historical time interval for which is evaluated; and
(b) the geographic distribution of the sources, as shown
Neutral stability
by Barone and Lovati (1980).
Wind speed" 6 m/s Moreover, the significance of this index is strongly
No rain reduced because the single-event mean mortality indices
distribution is rather fiat. As for other recorded toxic
Fig. 1. Total committed effective dose equivalent to individuals from releases, the most important is that of ammonia, which
(a) a large-scale accident release and (b) the revised estimate. Dose
from deposited activity are integrated over 30 yr. Adapted from has also occurred frequently in our century. The mean
Becher et al. (1981). mortality index in this case is 15 times smaller than the
index for chlorine. However, recent events (e.g., the
Seveso accident in 1976) clearly showed the inadequacy
of this approach for extremely toxic materials. In order
immediate health effects would thereby also be reduced. to have comparable acute effects for highly toxic chemi-
To conclude, it is particularly important that further cals, amounts considered are of the order of kilograms
data be obtained on iodine release retainment mecha- instead than tons. In order to get a comparative estimate
nism and containment performance in case of core melt- of risks for different factories that deal with different
down. toxic substances, the "chlorine equivalent" concept has
been introduced by the U.K. Health and Safety Com-
Chemical Plants mission (HSC, 1979). This is defined as the amount of a
The approach followed in the chemical sector for given substance which, when released in the atmosphere,
safety analysis is rather different from the nuclear one. produces the same number of fatalities of one ton of
It should be noted, indeed, that apriori analyses of acci- chlorine. As the effective value of the chlorine equiva-
dent consequences are rather unusual and recent (HSE, lent is derived (also taking into account the historical
1978; HSC, 1979; Ramsey et al., 1982; Pocchiari and evidence), it reflects not only the toxicity of a given
Silano, 1983; Coulston and Pocchiari, 1983). This kind substance, but also the characteristics of its atmospheri-
of analysis, moreover, neither routinely enters in the cal dispersion and the physical properties of the source.
licensing procedures nor are among the tools commonly On the basis of the chlorine equivalent concept the HSC
used by the chemical industry when investigating the report (1979) fixed the "inventories requiring notifica-
safety level of chemical plants. Only recently has a risk tion" where a 10-ton chlorine equivalent notification
assessment effort in the chemical industry been ini- level is recommended.
tiated, also by transferring models and methods that Annex III of the EEC Council Directive 82/501
480 G.C. Venuti et al.
chlorine case, as quoted in the HSC report (1979), 10 10 2 103 104 105
limiting ourselves to the period 1970-1980, one obtains
a probability of 10-3. However, no reliable historical Fatalities/X
data are available from which one could derive the be-
Fig. 2. Frequency of chlorine accidents involving fatalities. Adapted
havior of probability as a function of the number of from Simmons et al. (1974).
fatalities. One attempt towards an "apriort" estimate of
this kind has been made by Simmons et al. (1974). The
authors have studied the potential risk associated with WHO, 1981, 1982a), in addition to design or engineered
railroad chlorine shipments in the United States. The safety features.
potential risk was calculated for an accident involving The main objective of an emergency response plan is
the rupture of a tank car and the release of 90 tons of minimization or prevention of any possible adverse im-
liquid chlorine. By making suitable assumptions (pre- pact of the accident on population, ecosystem and,
sumably valid only for the United States) on the at- possibly, on property. Availability of adequate pre-
mospheric diffusion and the population density, the planned procedures and resources is essential for han-
results shown in Fig. 2 are obtained. dling emergencies. Then, even if it is not predictable
By referring only to the mean mortality index and by whether an actual accident will follow a predetermined
making the reasonable hypothesis that a large part of scenario, operative plans for a graded emergency re-
the chlorine inventory is released, an amount of chlorine sponse, based upon an analysis of potential accident
of the same order of magnitude as that considered in the spectrum, should be available.
EEC Annex could lead to an event with 5 fatalities. Criteria for accident analysis, already discussed
Although imprecise and questionable, this calculation above, also may be used to determine the size, type,
stresses the importance of developing suitable safety nature, location, time-dependence, and rate of danger-
goals, both for the probability of accidental events and ous materials release, as well as direct and synergistic ef-
for their consequences, also for chemical activities. fects. The analysis should also take into account the
consequences in terms of environmental contamination
and possible exposure of both onsite workers and offsite
Emergency ResponseSystems residents, as well as environmental effects in the absence
of any emergency response. Possible contamination
For facilities in which the physical possibility of flam- pathways should be carefully considered. In the case of
mable or toxic materials release is foreseeable, even as a chemical accidents, attention should be paid also to
consequence of a highly unlikely accident, emergency possible products of reactions occurring during or after
plans should be established (IAEA, 1978; CEC, 1982; the accident.
Chemicaland nuclearrisk management 481
Time scales associated with an emergency action ferent substrata must be examined (which may imply
following an accident may be classified as initial, early, different analytical procedures), adequate quality con-
intermediate, and late phases. The first two phases cor- trol must be organized, and so on. A sequential moni-
respond to prerelease and release propagation of the toring strategy also may be of help, in order to optimize
toxic or flammable cloud, respectively. In the early information obtained and to minimize the number of
phase, the risk is mainly due to direct exposure to and analyses required. A continuous mathematical-statisti-
contamination from the polluted cloud, while in the in- cal processing of data may be necessary to evaluate
termediate phase the risk is mainly connected with the available data and to design further monitoring steps.
exposure to contaminated environment through dif- Lastly, a careful estimation about the expected environ-
ferent routes. In the late phase, the risk is mainly due to mental fate of the released chemical is necessary to pre-
the widespread contamination of the territory (if any). vent its further dispersion. Physicochemical properties
All these phases should be characterized by remedial of the chemical involved, as well as some related mathe-
and rehabilitation action as appropriate. In the case of matical models and procedures (e.g., fugacity-con-
nuclear and chemical accidents, the time range scales for centration approach, partition analysis, physical trans-
the four phases are in the order, respectively, of 30 min port models, etc.), may be of help to this step.
to a few days, a few hours, a few days to a few weeks, In the case of the abovementioned TMI accident, the
and a few weeks to a few years. availability of aerial monitoring and sampling systems
Emergency action planners should identify resources, was extremely useful. The possibility of remote sensing
procedures and protective measures needed during the of large areas in a short time, and of the direct sampling
abovementioned phases, taking into account the possi- of contaminating clouds, makes such systems appropri-
ble accident analysis, the characteristics of the plant site ate emergency monitoring needs during nuclear acci-
and of its environment, the density and distribution of dents. The possibility of remote sensing procedures for
local population, the land use, and so on. The protective toxic chemical contamination monitoring is at present
measures may include sheltering, prophylaxis, control being studied by the WHO Regional Office for Europe,
of affected areas, personal protection methods and in cooperation with the Istituto Superiore di Sanit~t in
equipments, decontamination procedures for affected Italy.
individuals, evacuation procedures for people nearby, Basic parameters, to be considered either during the
medical care, social and economic care, diversion of planning phase or when managing actual accident con-
food and water supply, use of stored animal feed, sequences, are the "Intervention Levels." These levels
decontamination of areas, and further contamination represent a range of values of specific environmental
diffusion prevention. parameters (for example, the environmental range of a
As a primary need, the emergency response plan given toxic chemical or radiation, e.g., in ground or in
should ensure the availability of experts and institutions water, etc.) to which a prefixed set of protective
who will provide emergency management; once the acci- measures has to be put in practice. Levels of general in-
dent has occurred, these sources should also produce the terventions may also be established (for instance, when
data and measurement needed to assess foreseeable con- ground levels are higher than a prefixed value, the
sequences in real time for decision making. For nuclear evacuation of population groups may be advisable). In-
facilities, as a consequence of the Three Mile Island tervention levels may be of great importance in prevent-
(TMI) accident, a specific effort has been made in the ing and reducing risks and damages for human health
United States to develop an "Emergency Operating and the environment; they are not mandatory, but simply
Center" connected with computer networks. This guar- represent reference criteria for decision makers. They
antees the availability of the Atmospheric Release Ad- should be defined at a national and, possibly, interna-
visory Capability (ARAC) system (Sullivan, 1982) to- tional level.
gether with reliable up-to-date meteorological data, thus On the basis of the "a priorf' analysis of the possible
allowing rapid estimation of dispersion pathways of accidents and their expected consequences, and of a
contaminants released in the air, as well as offsite conse- comparison between estimated environmental con-
quences. In Italy, the ARIES system has been developed tamination levels and intervention levels, the "Emer-
by ENEA for the same purpose (CNEN, 1981). gency Planning Zones" (EPZ) may be defined. These
In the case of a nuclear accident, a fast measurement zones represent the areas where specific intervention
of the contamination level can be performed in the field levels are reached and specific protective measures are
by using portable monitoring devices; for more specific needed. Planners must identify the resources and pro-
measurements, more refined instrumentation and methods cedures needed to implement the adequate protective
are required. The determination of toxic chemical envi- measures relative to the various possible EPZ.
ronmental levels requires, as a rule, a large number of Feasibility analysis of emergency plans should always
chemical analyses, often to be performed in a limited be performed; in particular, this analysis may be one
period of time. Moreover, adequate analytical methods fundamental criterion to be used in licencing procedures
need often to be established in a short time; many dif- for new facilities and plants. Contingency planning
482 G.C. Venuti et al.
should be a nearly continuous and self-adjusting activity, Becher, P. E., Gjorup, H. L., Meide, A., Micheelsen, B., Thykier-
to maintain reliability of the response system, when Nielsen, S., Pejtersen, V. S., and Petersen, T. (1981) Calculation
of dose consequences of a hypothetical large accident at a nu-
processes, operations, and products change inside the clear power reactor. RISO-M-2299, RISO National Laboratory,
plant or when land use, demographic distribution, avail- Roskilde, Denmark.
able resources, or other relevant parameters change out- Commission of the European Communities (1982) Radiological pro-
tection criteria for controlling doses to the public in the event of
side the plant. Lastly, it must be noted that simulation accidental releases of radioactive material. V/5290/82 EN, CEC,
of possible accident scenarios is an unique tool for the Luxembourg.
testing efficiency of contingency plans and emergency CNEN/DISP (1977) Requisiti e criteri dei siti suscettibili di in-
sediamento di impianti nucleari. CNEN/DISP (77)2, National
response systems. Atomic Energy Committee, Rome, Italy.
CNEN/DISP (1981) Sistema ARIES. Primo rapporto di attivith. In-
Conclusions ternal Report DISP-AMB/133/81, National Atomic Energy Com-
mittee, Rome, Italy.
The source features for commercial nuclear reactors Coulston, F. and Pocchiari, F. (1983) Accidental Exposure to Diox-
ins. Human Health Aspects, Academic Press, New York, NY.
are well known through detailed studies that analyze the
European Economic Community (1982) Direttiva del Consiglio del
entire range of accidents physically possible, and the 24/6/82 sui rischi di incidenti rilevanti connessi con determinate
probability of each accident has been evaluated. As attivith industriali. L230, EEC, Brussels.
compared to earlier ones, recent assessment concerning Giorup, H. L., Micheelsen, B. and Thykier-Nielsen, S. (1982) Conse-
quences of large reactor accidents calculated on basis of empirical
more serious accidents seems to indicate lower proba- data, in Proceedings o f the International Conference on Nuclear
bilities of occurrence and lower rates of related accident Power Experience, Vienna, 13-17 September, vol. 4, pp. 119-130.
International Atomic Energy Agency, Vienna.
mortality. The methodology with which nuclear energy
Health and Safety Commission (1979) Advisory committee on major
safety problems are approached is rather well tested, hazard; Second report. Her Majesty's Stationery Office, London.
and it constitutes the basis for laws and regulations in Health and Safety Executive (1978) Canvey" An investigation of
force throughout the industrialized world. Apart from potential hazards from operations in the Canvey Island/Thurrok
area. Her Majesty's Stationery Office, London.
the current laws that reflect Euratom directives, Italy International Atomic Energy Agency (1979) Planning for off-site
also has statutory procedures that involve both national response to radiation accidents in nuclear facilities. IAEA-
TECDOC-225, IAEA, Vienna.
bodies and local authorities.
Technical Commission for Nuclear Safety and Health Protection
The methodology for analyzing chemical accident (1978) Rischio residuo di incidenti in centrali nucleari di tipo pro-
risks appears to be not so well established. Only during vato: Influenza sulla valutazione dei siti proposti e sui piani di
recent years have preliminary steps been taken towards emergenza. Study group report, Technical Commission, Rome,
Italy.
systematic risk assessment in the chemical industry, Marshall, V. C. (1982) Quantification as a mean of control of toxic
sometimes adopting the methods and models already hazards, in Proceedings of the symposium on the Assessment of
used in the nuclear industry and other fields. Partly Major Hazards", Manchester, 14-16 April, pp. 69-85. The Institu-
tion of Chemical Engineers, North Western Branch, London.
because of the vast variety of chemical plant types and Morewitz, H. A. (198 I) Fission product and aerosol behaviour follow-
of the substance involved, present-day risk studies are ing degraded core accidents, Nucl. TechnoL 53, 120-134.
based essentially on data collected after accidents and Nuclear Regulatory Commission (1975) Reactor safety study.
WASH-1400 (NUREG-75/014), NRC, Washington, DC.
on available toxicological data. The EEC Directive re- Nuclear Regulatory Commission (1983) Policy statements on safety
cently approved should act as a strong stimulus for the goals for the operation of nuclear power plants adopted by the
analysis of accidents associated with each high-risk Nuclear Regulatory Commission on January 10. NRC, Washing-
ton, DC.
plant and for a careful assessment of the impact on the Pocchiari, F. and Silano, V. (1983) Contingency planning for and
surrounding territory. Currently in Italy, statutory response to emergencies and accidents involving toxic chemicals,
responsibility for licencing chemical plants belongs Chemosphere 12, 743-748.
Ramsay, C. G., Sylvester-Evans, R., and English, M. A. (1982) Siting
largely to local authorities and the technical and scien- and layout of major hazardous installations, in Proceedings of the
tific bodies of the National Health Service act in an ad- symposium on the Assessment of Major Hazards, Manchester,
visory, or consultative, capacity. 14-16 April, pp. 335-351. The Institutions of Chemical Engineers,
North Western Branch, London.
The analysis and comparison of risks in chemical and Sampaolo, A. (1980) CEE direttiva Seveso, Sicurezza Sociale Oggi 2,
nuclear areas is still open. An appropriate culture for 57-66.
social and economic development would call for a Simmons, J. A., Erdmann, R. C., and Naft, N. B. (1974) The risk of
catastrophic spills of toxic chemicals. Internal Report UCLA-
choice of homogeneous health and environmental ob- ENG-7425, School of Engineering and Applied Science, University
jectives coming from a scientific, technological, and of California, Los Angeles, CA.
financial long-range effort. Sullivan, T. J. (1982) ARAC, VAX system design criteria. Report
UASG 82-1, Lawrence Livermore National Laboratory, Liver-
Acknowledgements-Stimulating discussions with G. Tenaglia of more, CA.
ENEA-DISP are gratefully acknowledged. World Health Organization (1981) Planning emergency response
system for chemical accidents. Administrative guidelines, WHO,
References Regional Office for Europe, Copenhagen, Denmark.
World Health Organization (1982a) Guideline document on
Barone, D. and Lovati, A. (1980) Considerazioni sulla stima della rehabilitation following chemical accidents. ICP/RCE 903 (20)/6
magnitudo delle conseguenze di un incidente. CNEN-RT/DISP bis, WHO, Copenhagen, Denmark.
(80)5, National Atomic Energy Committee, Rome, Italy.