TNAG-1776-FCO40-2530-Daya-Bay-nuclear-power-station-project-safety-concerns-in-Ho-1988 — Page 157

:

6.8 Duration of Incident

It is not possible to know before the event what the duration of the incident will be, hence it is not possible to forecast accurately the length of time the CCC, MAC and Press Briefing Centre will be required to be manned. As a guide, however, preparations should allow for the possibility of these centres being manned for 24 hours per day for at least 7 days. These preparations will include consideration of how the key posts will be manned throughout this period. This will involve training several people to fulfil each of the key posts.

6.9 Use of Computer Equipment

Atmospheric Dispersion Modelling

There are a number of computer codes available which may be used for predicting the downwind dispersion of radionuclides following a major reactor accident, eg the UMPIRE code (ref 19). Many of these codes are based upon two-dimensional Gaussian diffusion models which are at their most accurate when predicting effects at a few hundred metres to a few ten's of kilometres downwind from the source, and over fairly flat ground. Other more complicated models may use, for example, 'particle in a cell' methods coupled with wind field models to predict dispersion out to greater distances and over more difficult topography.

Daya Bay is a coastal site which is subject to diurnal sea breezes. The ground around the site is fairly hilly and as such the wind-fields around the site are likely to be complex. Detailed atmospheric dispersion models are likely to be difficult and costly to develop. Simple models, or the use of manual calculation techniques, for example as given in the NRPB Emergency Data Handbook (ref 20), are likely to be adequate for training purposes and for early hazard assessments following a release, provided the degree of potential error is known and allowed for.

The Royal Observatory have the necessary expertise and meteorological data in order to assess whether or not simple diffusion models would be adequate for accident assessment purposes. The Environmental Protection Department have expertise on atmospheric diffusion modelling. As such they should be jointly tasked with reviewing relevant models with a view to adopting a suitable method of calculating atmospheric dispersion, bearing in mind local meteorological and topographical effects.

One principal advantage of using computer codes, rather than manual techniques, is that it allows many radionuclides to be considered. If the release is predominantly iodine and noble gases this may not be important. In other circumstances the ability to take into account many radionuclides may be important.

Simple computer models can be valuable when preparing radiological data for exercises. Consideration should therefore be given to obtaining a suitable system for this purpose at least.

If computers are used to assess radiological consequences, incident assessors must also retain the ability to carry out manual calculations in case of computer failure.

Data Storage and Retrieval

Following a major incident large numbers of radiation measurements, and environmental sampling results will be generated. Consideration should be given as to how these data should be handled. A manual paper system may well be adequate provided it is well organised. Alternatively a small micro computer with a suitable database software package may be considered. If a computer-based system is to be used then the system should also be used for storage of routine environmental data to ensure that the operators are familiar with the system.

37