To evaluate the construction costs that, as said, represent the main component of nuclear LCOE, two approaches are usually adopted: top-down cost estimation and bottom-up cost estimation.

• Top-down estimation. The cost is calculated starting from a reference, known cost value, then considering the most important cost drivers that characterize the economics of that specific technology to derive scaled or proportional costs. Regarding the power plant industry, these drivers are: the plant size, the number of units to build, the site location, etc. This procedure is particularly appropriate when the plant design is still in the early phase of development, or when the plant design is characterized by a high level of complexity and number of systems as to make the cost estimation of each of them a hard task with a decrease in the reliability of the end result. An application of top-down cost estimation to SMRs is presented in Carelli et al. (2010).

• Bottom-up cost estimation. The cost analysis is carried out at ‘component level’ and the final cost is the sum of all of the costs related to the components manufacturing, assembling, operation, etc.

Once estimated through the above-mentioned procedures, the life-cycle costs, together with the cost of financing (equity and debt) and tax burdens, may be elaborated to perform a discounted cash flow (DCF) analysis. The DCF analysis provides the most relevant indicators of economic performance, such as the internal rate of return (IRR), the net present Value (NPV), the (LUEC) and the payback time (PBT) (see Figure 10.1). Several studies indicate that optimism in the cost estimation of large projects, such as civil and transport infrastructures, power plants, etc., is a common characteristic. This phenomenon may be observed in the case of NPPs, which are historically characterised by delay in construction and cost escalation (Locatelli and Mancini, 2012a). In order to provide a reliable cost estimation of SMRs, it is important to understand why the estimations of NPP costs, as well as large engineering projects, are usually so inaccurate and how to improve this process. Under this perspective,

Flyvbjerg et al. (2003) show that the availability and reliability of data about large projects affect the estimation. The authors identify two macro-categories of causes to explain inaccuracy in the cost forecast: (i) inadequacy of the methodologies and (ii) strategic data manipulation. The latter, combined with ‘optimism bias’, is responsible for most of the cost escalation.