2.4.2 It is useful to begin by identifying a 'long list' of options, containing all the initial ideas about possible solutions. This should include not only the conventional solutions, but also any more innovative suggestions, however outlandish they may at first appear. Imaginative thinking should be encouraged. Initial Agreements and Outline Business Cases should record all the long listed options.
2.4.3 Where the technical feasibility of one or more of the long-listed options is in doubt, and if there is a reasonable prospect that they will offer a VFM solution, it may be appropriate to commit resources to exploring further whether the option(s) can be delivered and at what cost. This may be considered, for instance, where particularly innovative solutions are in view. How much effort to devote to this is a matter for case-by-case judgement. Sometimes a separate 'feasibility study' report may be prepared. Such feasibility studies should be treated as inputs to the process of refining the options within an appraisal. They should not be confused with or regarded as a substitute for an appraisal or business case.
2.4.4 The long-listed options usually need to be sifted to produce a more manageable 'short list' of options for in-depth appraisal. This should be done according to specific, stated criteria. These may be expressed in terms of, for example, failure to satisfy the principal objectives of the proposal, or violation of important constraints regarding finance, manpower availability, policy commitments, and site suitability and so on. Where options are rejected in this way, the precise manner in which they fail to meet the criteria should always be explicitly explained.
2.4.5 Where approval is granted in stages, as per the requirements for business case approvals and delegated limits, it may be the case that some options can be ruled out at the preliminary option appraisal stage, so that the more detailed analysis is confined to a narrower range of options. However, this will not always be appropriate. In any event, good reasons should always be given for rejecting options.