MISTRAL V1.1.1: assessing doses from atmospheric releases in normal and off-normal conditions

Protecting the environment and the public from radioactive and chemical hazards has always been a top priority for all companies operating in the nuclear domain. In this scope, SGN provides all the services the nuclear industry needs in environmental studies especially in relation to the impact assessment in normal operating conditions and risk assessment in off-normal conditions. In order to quantify dose impact on members of the public due to atmospheric releases, COGEMA and SGN developed MISTRAL V1.1.1 code. Dose impact depends strongly on dispersion of radionuclides in atmosphere. The main parameters involved in dispersion characterization are wind velocity and direction, rain, diffusion conditions, coordinates of the point of observation and stack elevation. MISTRAL code implements DOURY and PASQUILL Gaussian plume models which are widely used in the scientific community. These models, applicable for distances of transfer ranging from 100 m up to 30 km, are used to calculate atmospheric concentration and deposit at different distances from the point of release. MISTRAL allows the use of different dose regulations or dose coefficient databases such as: - ICRP30 and ICPR71 for internal doses (inhalation, ingestion) - Despres/Kocher database or US-EPA Federal Guidance no.12 (ICPR72 for noble gases) for external exposure (from plume or ground). The initial instant of the release can be considered as the origin of time or a date format can be specified (could be useful in a crisis context). While the context is specified, the user define the meteorological conditions of the release. In normal operating mode (routine releases), the user gives the annual meteorological scheme. The data can be recorded in the MISTRAL meteorological database. In off-normal conditions mode, MISTRAL V1.1 allows the use of successive release stages for which the user gives the duration, the meteorological conditions, that is to say stability class, wind speed and direction and rainfall intensity. When the meteorology is specified, the user characterizes the release. In off-normal situation, successive stages can be defined independently from the meteorological stages. For each one, radionuclides are chosen, the activity released is given and the lung absorption type is specified in accordance with its chemical form. All those compositions can be recorded. In case of annual releases, release duration is automatically set to 1 year. When the release is fully specified, the receptors have to be characterized by giving their coordinates or their distance and angle from release point. Doses can be calculated at various moments specified by the user. In case of annual releases, the standard observation times correspond to one year but the user has the possibility to calculate for any whole number of years up to 70 years. Finally, the user has to specify the output data needed: primary data such as concentrations in air or on soil (at the different moments specified or integrated) or doses. For the public, five potential exposure pathways are available in MISTRAL V1.1.1. Code: - From plume: internal and external exposure, - From deposit: external exposure, internal exposure due to inhalation of resuspended radionuclides and ingestion. Exposure from plume occurs as soon as radionuclides are released in atmosphere whereas deposited activity generates long time exposure. Dose impact is calculated at all distances and times stated by the user for different age classes. Ingestion is calculated by using a dynamic food chain model. Ingestion dose is calculated by using annual food consumptions. The user has to possibility to record its own data. In off-normal situations, doses from deposit are integrated over 30 days or one year duration. If the routine release scheme is chosen, doses are integrated over the duration specified by the user as observation times. The time spent inside and outside by any individual can be specified. In addition, attenuation coefficient for internal and external doses can be given. MISTRAL has been validated through comparisons with experimental data and other codes such as SIROCCO, COTRAM (using DOURY) or ALOHA (using PASQUILL). At SGN, MISTRAL is systematically used in projects relative to French companies (COGEMA, CEA and ANDRA) or foreign parts (YUCCA MOUNTAIN (USA), ENRESA (Spain), Chernobyl Intermediate Storage Facility (Ukraine))