SEMER: a simple code for the economic evaluation of nuclear and fossil energy-based power production systems

The code system, SEMER, was recently developed to evaluate the economic impact of various nuclear reactors and associated innovations. Models for nearly all fossil energy-based systems were also included to provide a basis for cost comparisons.Essentially, SEMER includes three types of model libraries: the global model, for a rapid estimation of various nuclear and fossil energy-based systems, the detailed models, for the finer cost evaluation of individual components and circuits in a PWR type of reactor and the fuel cycle models, for PWRS, HTRs and FBRs, allowing the cost estimations related to all the steps in the nuclear fuel cycle, including reprocessing and disposal.This paper summarises our on-going investigations on new developments in, and on the validation of, the SEMER system.Details of the modelling principles, and the results of validation carried out in the context of an EDF/CEA Joint Protocol Agreement, are also presented.First results of this validation are highly encouraging:

• Relative errors for the total kWh or overnight and investment costs are less than 5% for large PWR systems operating in France or other countries.

• These errors are less than 3% for small-sized compact PWRs and they are of the order of 4–7% for HTRs (as compared to IAEA estimations).

• For fossil energy-based power plants, the relative error, even with slightly different cost breakdown between SEMER and that of existing installations, is from 5 to 20%.

• Similarly, errors on the nuclear fuel cycle costs are about 1–4%, compared to published reference values.

Article Outline

1. Introduction

2. The models

2.1. The global models

2.2. The detailed models

2.3. The fuel cycle model

3. Cost modelling principles

3.1. Input data and output

3.1.1. Input data

3.1.2. Output

3.1.3. Interest during construction

3.2. An illustrative example of power cost calculations

4. The fuel cycle model

4.1. An illustrative example of fuel cycle calculations

5. Validation

5.1. Validation results for nuclear reactors

5.2. More recent validation of operating power plants

5.3. Circuits, tubes and components

5.4. Fuel cycle costs comparisons

6. Conclusions

References

1. Introduction

This paper describes some of the salient features of the economic evaluation models, integrated in CEA’s code system, SEMER (Système d’Evaluation et de Modélisation Economique de Réacteurs).The basic aim of this development is to furnish top management and project leaders a simple tool for cost evaluations enabling the choice of competitive technological options.In the particular context of CEA’s R&D innovative programme, it was imperative to include this economic dimension in order to assess the economic interest of the proposed innovations and to search for other promising areas of R&D, leading to nuclear power cost reductions.SEMER is actually used in the form of a totally machine-independent and user friendly interface in the JAVA language.

2. The models

There are three distinct categories of models in the SEMER system.

2.1. The global models

These models are designed for a quick overall economic estimation. Current version of SEMER includes models for:

• Nuclear power plants, such as PWR of the 1400 MWe type (double confinement and four loops), PWR of the 900 MWe type (single confinement, three loops), HTGR (high temperature, gas-cooled reactor), LTR (integral nuclear reactor for heat production), NP (compact PWR) and PWR-C (modular integral PWR such as the SIR concept).

• Conventional, fossil energy-based power plants, such as pulverised (or fluidised bed), coal-fired plant, with desulphurisation treatment, oil-fired plant, gas-fired plant and diesel plants of all types. Also included are gas turbine plants, plant with a simple gas turbine, plant with a combined cycle gas turbine (“indoor” and “outdoor” constructions).

2.2. The detailed models

This option allows detailed cost estimations by individual modelling of reactor components, circuits and associated buildings, etc. In the present version, only the following models for PWR are available:

• Reactor components, such as civil engineering of associated buildings and structures, reactor vessel, steam generator with U-tubes, steam generator with straight tubes, the pressuriser, primary circuit pumps, the travelling crane, cooling tower, cooling tower with mechanical ventilation, turbine-driven pumps, pump motors, centrifugal pumps, air ejectors, heat exchanger casing, special tubes in stainless steel and special tubes in black steel, with internal coating in stainless steel.

• Reactor circuits, including: (1) basic circuits, such as primary circuit connecting the core, pressuriser, primary pumps and steam generator and secondary circuit connecting the steam generators and turbines; and (2) auxiliary circuits, such as steam generators blow-off circuit, steam generator emergency feed-water circuit, confinement spray system, chemical and volumetric control system, emergency core cooling system, component cooling system, water make-up and boron circuit, nuclear sampling system, drain, vent and exhaust circuits, residual heat removal system, effluent control and rejection system and diverse other circuits inside and outside the reactor building.

For the economic evaluation of an innovative PWR, the detailed models allow to take into account the specificities of the new concept and thus bring corrections to the global model, available in the SEMER library and considered having the closest analogies to the innovative PWR to be evaluated. This approach was used in Nisan et al. (2002) to evaluate the AP-600.

2.3. The fuel cycle model

In addition to the above, SEMER also incorporates a detailed model for the fuel cycle cost calculations of a nuclear reactor, treating all the stages of the nuclear fuel cycle from ore extraction to ultimate disposal, with the following options:

• Uranium oxide (UOX) cores.

• 100% mixed, uranium–plutonium oxide (MOX) cores.

• Cores with first loading in UOX, then equilibrium core in MOX.

• Mixed cores with x% MOX fuelled assemblies (under development).

• HTR cores and fast reactor (EFR type) cores.

Several options regarding the treatment of the fuel cycle front- and back-ends are also available:

• Global treatment as in the IAEA WREBUS study (IAEA, 1992).

• Detailed treatment as in the OECD study (OECD, 1994). This is the default option.

• A combination of the above, with a semi-detailed calculations, including the specific treatment and costs for B and C type of wastes, as used by the French Ministry of Industry, DIGEC and by EDF (DIGEC, 1997).

• The CEA model, derived from feed-back of experience for front- and back-end operations.

It should be noted that the standard OECD option includes all the steps in the fuel cycle from the mine to final disposal. The WREBUS option only considers a global value for the fuel cycle back-end. The EDF model (detailed in Table 10) is in between. Finally, in the CEA model, all the costs concerning the front-end, the fabrication and enrichment and the back-end (reprocessing, then final disposal) are expressed as polynomial expressions derived from the costs of a large number of real cases.

3. Cost modelling principles

The basic principle governing the development of models in the SEMER system is the fact that, for most projects, especially in their preliminary phases, it is sufficient to first make a relative cost estimation by the simplest and fastest methods available. The results obtained are then further refined in the final stages of the project when relevant choices of options and technologies are almost fixed. The only condition is that consistent estimating techniques be used so that alternatives can be compared on the same basis, and comparisons can also be made between competing projects.This principle was first used in the chemical and petrochemical industries where continued development over several decades has produced simple but powerful methods for cost evaluations (Popper, 1970).These methods were adapted to nuclear reactors and further developed at CEA during the last 20 years. They have been successfully applied, in particular for the cost assessment of nuclear submarine reactors, operating large-sized PWRs, new small- and medium-sized reactor concepts as well as for a variety of technologies and components, utilising nuclear or fossil energies.The basic steps involved in the development of such methods are:

1. The power plant cost is first carefully decomposed into several “cost modules”. This method was first proposed in the early 1970s for chemical plant cost estimations (Guthrie and Grace, 1970). An estimating module represents a group of cost elements (or items) having similar characteristics and relationships. Each of these elements can be made to represent a given function in the overall module (e.g. site acquisition and development, major process equipment such as a heat exchanger, a pressure vessel, etc.).

2. A detailed study is then made to make an inventory of the various generic models¹ which bear a sufficient number of analogies with the module that one would like to assess. Thus, for example, the cost evaluation model for the PWR pressure vessel was developed from the available models for the stainless steel lined high pressure reservoirs used in the industry.

3. The cost C_i of an element i in a given module is then mathematically expressed in the form of simple equations of the type:

(1)

C_i=A_i+(B_i×P_iⁿ)

where A, B and n are the so-called “adjustment coefficients” and P is power or capacity (electric power of a reactor, for example).

4. The adjustment coefficients are then obtained by applying well-known mathematical techniques (a least-squares fit of a data base, for example) for a large number of values for P.

5. To qualify the algorithms, developed as above, the models are more finely tuned from the results of published data, taking into account the use of field materials, field labour and other industrial factors.

6. Finally, a validation of the model is undertaken by comparison with the “real” values from existing installations.

The SEMER system was basically developed for the assessment of innovations in reactor systems, made in the context of the French Nuclear Power Programme. The adjustment coefficients were then obtained from available data bases for experimental, operating or nuclear submarine PWRs and the fossil energy-based electricity producing systems. This is the main reason that the basic costs of most items need to be expressed in French Francs (FF) which are then converted into Euros or US dollars. Some information on other reactor types, e.g. HTRs, was also obtained from external sources such as the IAEA. In its current form, SEMER remains nonetheless highly oriented towards PWR type of technology.However, because of the inherent generic nature of the built in models, they can be easily adapted to treat other reactor systems. One could, for example, use the model for combined cycle gas turbines, to develop part of the models for HTRs with direct cycles.

3.1. Input data and output

3.1.1. Input data

Efforts were made to harmonise the input and output data for all power plant types, with only minor and easily comprehensible modifications in the input data.Examples of input data panels, for the global models of a nuclear reactor and a fossil fuelled plant, are summarised in Table 1.