Conduct of Operations

DOE-STD-6003-96

V&V process should be documented. Documentation should be prepared to manage the config-

uration control of the software itself.

Many of the standards and requirements used to verify and validate computer codes were

developed for commercial nuclear power plants. Guidance information is embodied in ASME

NQA-1 and ASME NQA-2 standards, as well as several American National Standards Institute

(ANSI)/Institute of Electrical and Electronic Engineers (IEEE) Standards (ANSI/IEEE STD 730,

828, 829, 830, 983, and 1012).

Verification is defined as the process of determining whether the software is coded cor-

rectly and conforms to the specified software requirements. Full verification would require a line-

by-line check of the entire computer code to ensure correctness. However, other less stringent

methods are considered applicable, such as developing a series of calculational cases or input

decks that test much of the logic in the code to ensure that the code performs as stated in the

users' manual. As a general rule, design and safety analysis should be verified because it is

good engineering practice.

Validation is defined as the process of evaluating software to ensure compliance with

software requirements and physical applicability to the process being modeled on the hardware

being used. Validation is generally more involved than verification. Validation of a code consists

of comparing its output with known analytical solutions for problems similar, yet perhaps simpler,

than the problem at hand. Validation also includes benchmarking the code against relevant

experimental data, thus ensuring that the analysis reasonably captures the correct physics and

chemistry. Validation can also include comparison with an existing, already validated, computer

code.

The number and type of benchmarking problems needed to validate a computer code are

functions of the complexity of the phenomena being modeled, the codes range of applicability,

and the data that are or could be available. For a complicated computer code, verification could

require that individual models and submodels in the code be V&V using separate-effects data

and that integral validation of the code also be performed. These issues are functions of the

specific technical areas and need to be considered in the respective V&V processes.

Due to the current experimental nature of fusion devices, it may not be possible to com-

pletely verify and validate a code. In such cases, other options should be explored to assure

safety of the facility. These options may include but are not limited to the use of test coupons to

be evaluated after specific periods of operation and qualification of materials/equipment using

deuterium-deuterium operations before using tritium as a fuel.

4.4.4 Conduct of Operations

Experience has shown that the better operating facilities have well-defined, effectively

administered policies and programs to govern the activities of the operating organization, includ-

ing the areas described by these guidelines. The guidance is based upon well-developed indus-

trial operations practices. They are written to be flexible, so that they encompass the range of

facilities and operations.