2015 International Nuclear Atlantic Conference - INAC 2015 São Paulo, SP, Brazil, October 4-9, 2015 ASSOCIAÇÃO BRASILEIRA DE ENERGIA NUCLEAR - ABEN ISBN: 978-85-99141-06-9
SELECTION OF RELEVANT ITEMS FOR DECOMMISSIONING COSTING ESTIMATION OF A PWR USING FUZZY LOGIC Deiglys Borges Monteiro1, Alexander Lucas Busse2, João M.L. Moreira 3 and José Rubens Maiorino4 Centro de Engenharia, Modelagem e Ciências Aplicadas – CECS, Programa de Pós-Graduação em Energia e Engenharia da Energia Universidade Federal do ABC Campus Santo André, Avenida dos Estados, 5001, Bangú 09210-580 Santo André, SP 1
[email protected] 2
[email protected] 3 joao.moreira @ufabc.edu.br 4
[email protected]
ABSTRACT The decommissioning is an important part of a nuclear power plant life cycle which may occur by technical, economical or safety reasons. Decommissioning requires carrying out a large number of tasks that should be planned in advance, involves cost evaluations, preparation of plans of activity and actual operational actions. Despite the large number of tasks, only part of them is relevant for cost estimation purpose. The technical literature and international regulatory agencies suggest a variety of methods for decommissioning cost estimation. Most of them require a very detailed knowledge of the plant and data available suitable for plants that are starting their decommissioning but not for those in the planning stage. The present work aims to apply fuzzy logic to sort out relevant items to cost estimation in order to reduce the work effort involved. The scheme uses parametric equations for specific cost items, and is applied to specific parts of the process of nuclear power plant decommissioning.
1. INTRODUCTION The decommissioning of Nuclear Power Plants (NPPs) is an important part of its life cycle in which management and operational tasks are employed [1],[2] aiming to reducing risks inherent to its operations and other activities, avoiding the spread of radionuclides [1]-[4]. At the end, the site should reach the final state planned before decommissioning begins and the regulatory licenses are finished [5],[6]. The decontamination and dismantling tasks shows the highest difficulty, risks and costs among other decommissioning tasks [7]. The decommissioning could be conducted according different strategies, differing between each other in terms of the time need, tasks organization, occupational and environmental risks and associated costs. The most known strategy are, namely: DECON (Immediate Dismantling), SAFSTOR (Postponed Dismantling) e ENTOMB (Entombment) [5],[8]-[12]. DECON strategy allows to release site from license control effort in less time at the expense of greatest occupational risks, difficulty in tasks execution and higher costs [5],[8]. In SAFSTOR strategy, the time need between the begining and the end of decommissioning varies from a couple of years to decades, tipically 30 years, according the final state desired to the site, accepted occupational risks and time need to site clearance from licences
[5],[8],[13],[14]. Finally, ENTOMB strategy usually shown the highest time lenght, overcoming institutional time control (over 300 years), requiring in this way the construction of confinement structures using materials resistant to long term requests, such as weather, entombing in this way the contaminated and activated parts for the time need to they do not pose unacceptable risks anymore [5],[8],[13],[14]. In SAFSTOR and ENTOMB the NPP should be shutdown and kept in a safe state, allowing the decay of activated parts to a level that permits it could be safely dismantled. As consequence, costs with security and maintanence of NPP facilities will raise, while decontamination and dismantling costs will be reduced [5],[8],[13],[14]. In any case, each strategy could be conducted in different manners. That is, NPP systems could be considered fully or divided in its individual components [15]. The choice for one or other approach frequently is aided by virtual mock-ups, which lets planners to visualize NPP rooms or places with higher radiological activity, preventing workers long term exposure to high radiological doses and other occupational risks [16]-[18]. The decommissioning operational tasks have as consequence the production of a wide range of wastes, commonly sort according its radiological activity. They are produced mainly by decontamination and dismantling operations (D-D). The properties and other features of these wastes changes in function of D-D techniques employed, which should be segregated, treated, packaged and stored (temporarily features) or disposed (permanently features) [1],[2],[19]-[22]. Once the decommissioning begins, it should be assured that there will be enough financial resources to cost all it. In this manner, regulatory agencies requires financial guarantees. Thus, an estimation of costs involved is necessary to plan the most suitable strategy to raise funds. The solution frequently is composed by some methods even if the most adopted method are insurance payments and charging collection together the energy sale [1],[5],[8],[11],[12]. The cost estimation could be employed according different methods. However, actually the method proposed by ISDC [23] (International Structure for Decommissioning Costing) is the method suggested by international regulatory agencies [1],[3],[5]. The ISDC [23] presents a flexible structure, allowing apply it with different detail degree as need. According ISDC [23], the decommissioning costing accuracy could range from -30% to 50% for estimative, from -15% to 30% for budgets costs and from -5% to 15% for final costing. In order to define the best solution to raise funds to decommissioning cost, an estimative of costs involved should be conducted during the NPP operations or, in the best case, before beginning of its construction and operation. Due to the greatest variable number which could affect the final value, a strong effort in data collection should be employed, which could results in time increasing to it [24]-[26]. In this way and considering the error margin accepted for estimative costing, only some relevant NPP systems and equipments items should be take into account. To help the selection of them, a computational code could be developed, constituting as an important planning tool. The Fuzzy logic is a technique which infer results in a simple way using computational codes which, in turn, uses data entered by an user. The obtained result could be shown as different scenarios, allowing different approaches to the problem with the same final result. The input data could be sort in categories or matrices, reducing the complexity of them [17]. Briefly, the Fuzzy logic method is divided in 5 steps: data input, processing or data encoding (Fuzzyfication), metering or inference, decoding (DeFuzzyfication) and results [17]. The
INAC 2015, São Paulo, SP, Brazil.
Fuzzy logic could be used in a wide range of applications, including risks evaluation during nuclear reactor decommissioning [7]. The present work aims to evaluate the feasibility of application of Fuzzy logic to selection of relevant items during decommissioning for costing purposes from a section of the primary circuit in a PWR (Pressurized Water Reactor ) NPP using data from a reference plant. Once selected, these items could be used in a cost estimation to that section and to help in definition of the best strategy to raise funds for costing the decommissioning. The relevance of input data would be evaluated considering its final storage and transportation costs, which would be sort in 3 different scenarios. These scenarios, in turn, would be sort in function of its significance. 2. METHODOLOGY Schematically, the computational code under development, named Separador Fuzzy, has the structure presented in Figure 1.
Input data
User decision
Class A
Class B
Class C
Class D
Class E
Separador Fuzzy
Group 1
Group 2
Group 3
Group 4
Group 5
Group 6
Figure 1: “Separador Fuzzy” computational code structure The code Separador Fuzzy need normalized input data in terms of Waste final storage costs (α) and Waste transport to final repository costs (β), both in terms of monetary units by mass (units/kg). In addition, the code need input data of Total volume and Activity. All of them is necessary for the 5 material classes considered, which are composed by the follow equipment or plant sections, namely: A) Reactor Pressure Vessel (RPV) and its internals; B) Concrete biological shield dismantling; C) Electrical equipment removal; D) Primary circuit dismantling; E) Concrete structures dismantling. Table 1 presents for each category the normalized constants α e β obtained from a reference 1200MWe power PWR NPP. Thus, it is necessary that user has the data shown in Table 2 to normalize in percentual factors each item according it total value. Table 1: Normalized constants for each class Constants Class A – RPV and its internals Class B – Concrete biological shield dismantling Class C – Electrical equipment removal Class D – Primary circuit dismantling Class E – Concrete structures dismantling
INAC 2015, São Paulo, SP, Brazil.
α – Waste final storage cost 32,00 units/kg 4,59 units/kg 43,95 units/ m3 46,1 units/kg 4,59 units/kg
β – Waste disposal transport cost 2,62 units/kg 1,68 units/kg 43,95 units/ m3 0,67 units/kg 1,68 units/kg
Fonte: Valrey e Rusch (2001) Table 2 – Separador Fuzzy input factors description Parameter PTi VTi ATi
Description Total mass of wastes from ith class Total volume of wastes from ith class Total activity of wastes from ith class
Remark In any case for ith being each category (A,B,C,D e E)
The code Separador Fuzzy would sort normalized input data using adhesion functions as shown in Figure 2 in the following manner: "Pouco Relevante" (PR – Minor Relevance), presented in red, left side with value different of zero in Figure 2, or "Grande Relevância" (GR – Major Relevance), presented in black, right side with value different of zero in Figure 2. The processing would be done using the 15 rules shown in Table 3, which also relates them with the output groups, these last presented in Table 4.
Figure 2 – Adhesion functions of code “Separador Fuzzy” Table 3 – “Separador Fuzzy” processing rules Nº 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Rules If GR (Major Relevance) in Storage cost, then Group 1 is "Muito Pertinente" (MP – Very Pertinent). If GR (Major Relevance) in Transport cost, then Group 2 is "Muito Pertinente" (MP – Very Pertinent). If GR (Major Relevance) in Storage and Transport costs at the same time, then Group 3 is "Muito Pertinente" (MP – Very Pertinent). If GR (Major Relevance) in Volume and Activity at the same time, then Group 4 is “Muito Pertinente” (MP – Very Pertinent). If GR (Major Relevance) in Volume and in Storage cost at the same time, then Group 5 is “Muito Pertinente” (MP – Very Pertinent). If PR (Minor Relevance) in Storage cost, then Group 6 is "Muito Pertinente" (MP – Very Pertinent). If PR (Minor Relevance) in Transport cost, then Group 6 is "Muito Pertinente" (MP – Very Pertinent). If PR (Minor Relevance) in Volume, then Group 6 is "Muito Pertinente" (MP – Very Pertinent). If PR (Minor Relevance) in Activity, then Group 6 is "Muito Pertinente" (MP – Very Pertinent). If GR (Major Relevance) in Storage cost and PR (Minor Relevance) in Transport cost, then Group 3 is “Pertinente” (P – Pertinent). If PR (Minor Relevance) in Storage cost and GR (Major Relevance) in Transport cost, then Group 3 is “Pertinente” (P – Pertinent). If GR (Major Relevance) in Volume and PR (Minor Relevance) in Activity, then Group 4 is “Pertinente” (P – Pertinent). If PR (Minor Relevance) in Volume and GR (Major Relevance) in Activity, then Group 4 is “Pertinente” (P – Pertinent). If GR (Major Relevance) in Volume and PR (Minor Relevance) Storage cost, then Group 5 is “Pertinente” (P – Pertinent). If PR (Minor Relevance) in Volume and GR (Major Relevance) in Storage, then Group 5 is “Pertinente” (P – Pertinent).
INAC 2015, São Paulo, SP, Brazil.
Table 4 – Separador Fuzzy Groups description Group 1 2 3 4 5 6
Description Major Relevance (Grande relevância – GR) in Storage cost Major Relevance (Grande relevância – GR) in Transport cost Major Relevance (Grande relevância – GR) in Storage and Transport costs Major Relevance (Grande relevância – GR) in Volume and Activity Major Relevance (Grande relevância – GR) in Volume and Storage cost Minor Relevance (Pouco relevante – PR)
As result, the code would sort all items of some class according the affinity of their groups. The code output functions are presented in Figure 3, while the results are presented in 6 groups according Table 4. It should be highlighted that some items could be sort in several groups at the same time. The Table 5 presents the parametric equations developed and used by “Separador Fuzzy” code to sort the groups as presented in Table 4.
Figure 3: Separador Fuzzy code output functions Table 5 – Separador Fuzzy code parametric equations Equations
INAC 2015, São Paulo, SP, Brazil.
Description GR in Storage cost
(1)
PR in Storage cost
(2)
GR in Transport cost
(3)
PR in Transport cost
(4)
GR in Volume
(5)
PR in Volume
(6)
GR in Activity
(7)
PR in Activity
(8)
where the variables “P”, “V” and “A” of equations (1)-(8) means mass, volume and activity of each item, respectively. It was assumed that only items which has values greater or equal 1% of the total for that item, was assumed as Major Relevance (GR – Grande Relevância), considering a specific class (A to E), while minor values was assumed as Minor Relevance. The reason to adoption of this value is because, with exception of voluminous materials, case of concrete, all other presents a percentage value less or about 1%. The use of a higher value (higher than 1%) would result in a sort with most part of materials as Minor Relevance (PR – Pouca Relevância) unless that and β values were changed. Note that, among the 3 variables (Volume, Mass and Activity), for the Storage cost and Transport cost the most important variable is the Volume. After sort the input data according its relevance, the code proceeds with another sort in terms of pertinence as follows: Less Pertinent (Pouco Pertinente) for 0-1%, Pertinent (Pertinente) for 0,5-1,5% and Very Pertinent (Muito Pertinente) for 1-100%. Since the code uses all 5 classes data, some items could be sort as Major Relevance for his own class but, when this class is compared with the others, it could lose its importance, that is, this class could be sort as Less Pertinent. Thus, following the presented sort methodology, only items sort as Major Relevance of a class sort as Very Pertinent or, at least, Pertinent, would be take into account in the estimative decommissioning cost. In addition of the presented, the other methods used by Fuzzy logic in the code is the Mandani Method, DeFuzzification Method Centroid and a Triangular adhesion function, resulting in the previous shown Figures 2 and 3. 3. CASE STUDY – TROJAN NUCLEAR POWER REACTOR To check the feasibility of Fuzzy logic application to select only the relevant items of a NPP to be included in an estimative cost for decommissioning purpose, the input data was obtained from a reference NPP, actually under the finishing of its decommissioning process. Specifically, the reference plant choose was take as reference plant also in several reports about decommissioning by the American nuclear regulatory agency, the NRC (U.S. Nuclear Regulatory Commission). This plant is commonly known as TROJAN NPP [28]. Notwithstanding, this plant was choose since its nuclear reactor and nuclear steam supply systems constructive design are very similar to those of Brazilian NPP Angra 1, Angra 2 and in the future, Angra 3 and others whose could be constructed if the national energetic plan for 2030 year (Planejamento Energético Nacional 2030 – PEN 2030) take place [29],[30]. Finally, this choose was driven by the availability of reports and data of the NPP with the need detail level. Beyond the assumptions made, the volumes data was used also to approximate the masses, considering to this only 2 materials and its respective densities: for contaminated metals, since most part of them are composed by ferrous alloys (including stainless steel), it was adopted the steel density (7870 kg/m³) while for contaminated structures, since most part of them is composed by concrete, it was adopted the concrete density (2500 kg/m³). The TROJAN NPP data are available in Tables 6 and 7. The data used refers to nuclear reactor
INAC 2015, São Paulo, SP, Brazil.
primary system only. To run the code, it was choose randomly the follow equipment: “Cooling pumps and electrical motor drivers”. Table 6 – TROJAN reactor features Class
Volume [m3] 236,2 NA 1703,0 2060,5 1697,5
Mass [kg] 1.858.894 NA 16.216.135 4.243.750
RPV and internals Concrete biological shield dismantling Electrical equipment removal Primary circuit dismantling Concrete structures dismantling NA = Not available
Activity [GBq] 1,54.108 NA 37 1,6.104 37
Source: PGE-1061 (2001) Table 7 – Large TROJAN PWR primary circuit equipments radioactive wastes Systems Steam generator and Pressurizer (large components removed) Pipes Steam generator system Cooling pumps and electrical motor drivers Sensors, instrumentation and control rods mechanism Pressurizer relief tank RPV system Total
Volume (m3) 1636,7 166,9 100,9 86,2 48,9 17,7 3,3 2060,5
Activity (GBq) 8177 37 4958 3071 37 259 1,6x104
Source: PGE-1061 (2001) To the choose equipments, pertaining to Class D, it was developed parametric factors, as given by Table 8, considering the dismantling of these items. To other classes, these factors should be properly developed, not being object study of the present work. The expected result is that the code could be able to sort by relevance items pertaining to Class D, which should be sort according its pertinence. It should be observed that since data about Concrete biological shield dismantling were not available, in the occasion of data collection, they were not considered in the runs of the code. Table 8 – Class D parametric factors for TROJAN NPP Parametrics
Description Major relevance in Storage cost Minor relevance in Storage cost Major relevance in Transport cost Minor relevance in Transport cost Major relevance in Volume Minor relevance in Volume Major relevance in Activity Minor relevance in Activity
INAC 2015, São Paulo, SP, Brazil.
4. DISCUSSION AND RESULTS After run the code, it was obtained the results as shown in Table 9, sort according Table 4 groups. Table 9 – Output results Group
Description
Scenario
1
Major relevance in Storage
Very Pertinent
2
Major relevance in Transport cost
Less Pertinent
3
Major relevance in Storage and Transport costs
Less Pertinent
4
Major relevance in Volume and Activity
Less Pertinent
5
Major relevance in Volume and Storage cost
Less Pertinent
6
Minor relevance
Less Pertinent
Considering the results shown in Table 9, the equipments “Cooling pumps and electrical motor drivers” are Pertinent in the view of Storage costs only. However, considering the other classes, even Class D (dismantling tasks only) being sort as Major Relevance, the as mentioned equipments was sort as Less Pertinent. Thus, with exception of Storage costs, these equipments should not been considered in estimative costs. Notwithstanding, all data sort in the same way could have an alternative approach. This alternative approach could sum the contribution of all of them in an item called “Others”, which could be treated rightly after the main equipments were considered, in the same way (relevance and pertinence) or using another approach. This approach leads to a result with a minor error. In the view of the results presented by the proposed code using Fuzzy logic to select relevant items to decommissioning for costing estimative of a NPP, the code are considered able to be used as a planning tool with other tools commonly used. Despite the code is considered able to do the selection, the results should be validated, which are not done in this work at the present development stage of the code. The validation could be employed using the results of the code, with the sum of the selected items. Then, the sum could be compared with estimated cost available for a reference plant (as case of TROJAN NPP). The validation would enable the use of the code to select other relevant items from other NPP systems. It must be remembered that to apply the code to other systems, new parametric equations should be developed as well as the cost factors should need some correction. In addition, the proposed code was developed for PWR NPP. If the plant is of other type, the code should be adapted. 5. CONCLUSIONS It was presented a computational code using the Fuzzy logic to select relevant items in the decommissioning of a NPP for costing purposes. The code was developed for a 1200MWe PWR NPP or other with similar size and constructive type. Parametric equations was developed for a class of tasks and for a specific part of the NPP, both choose in the view of the data available to permit this. The tasks reefers to the dismantling of the primary circuit of the NPP and the part reefers to the cooling pumps and the electrical motor drivers. After data input and evaluation of output results, the code was considered able to do the purposed
INAC 2015, São Paulo, SP, Brazil.
objective, sorting in terms of relevance and pertinence the selected item. No validation was carried on in the present work. In addition, the developed parametric equations and the cost factors used needs to be changed to be applied to other systems and parts of the NPP. Notwithstanding, the code was developed for a 1200MWe PWR NPP only, and should need some changes to be applied to other size and construction type NPPs. The data used was obtained from a reference NPP, TROJAN. REFERENCES 1. D.B. MONTEIRO, J.M.L MOREIRA, J.R. MAIORINO, “Metodologia e equações para estimar o custo de descomissionamento de plantas nucleares brasileiras”, Anais do IX Congresso Brasileiro de Planejamento Energético, Florianópolis, Brazil (2014). 2. Matej ZACHAR, Vladimír DANIŠKA, Vladimír NEČAS, “Improved analytical methodology for calculation assessment of material parameters in nuclear installation decommissioning process”. Progress in Nuclear Energy, v.53, p.463-470 (2011). 3. Kwan Seong JEONG, Dong Gyu LEE, Chong Hun JUNG, Kune Woo LEE, “Structures and elements for the decommissioning cost estimations of nuclear research reactors”, Annals of Nuclear Energy, v.34, p.326-332 (2007). 4. IAEA, A Proposed Standardized List of Items for Costing Purposes, Interim Technical Document, OECD/NEA, IAEA, EC (1999). 5. G.J. KONZEK, R.I. SMITH, M.C. BIERSCHBACH, P. N. MCDUFFIE, “Revised Analyses of Decommissioning for Reference Pressurized Water Reactor Power Station”, Pacific Northeast Laboratory, Division of Regulatory Application, U.S.NRC, NUREG/CR5884, NRC (1995). 6. UFABC, Caracterização final do sítio da CNAAA, CECS Programa de Pós-graduação em Energia e Engenharia de Energia, Universidade Federal do ABC, UFABC-DESCOM-EST013-01, UFABC (2014a). 7. Kwan-Seong JEONG, Byung-Seon CHOI, Jei-Kwon MOON, Dong-Jun HYUN, JongHwan LEE, Geun-Ho KIM, Ho-Sang HWANG, Seong-Young JEONG, Jung-Jun LEE, “Risk reduction approach to decommissioning hazards of nuclear facilities”, Annals of Nuclear Energy, v.63, p.382-386 (2014). 8. N. G. WITTENBROK, “Technology, Safety and Costs of Decommissioning Nuclear Reactors At Multiple-Reactor Stations”, Division of Engineering Technology, U.S.NRC; NUREG/CR-1755, NRC (1982). 9. R. BARDTENSCHLAGER, D. BOTTGER, A. GASCH, N. MAJOHR, “Decommissioning of Light-Water Reactor Nuclear Power Plants”, Nuclear Engineering and Design, v.45, p. 1-51 (1978). 10. B.L. BAUMMAN, “Evaluation of nuclear facility decommissioning projects programstatus”, Nuclear Engineering and Design, v.89, p.47-50 (1985).
INAC 2015, São Paulo, SP, Brazil.
11. Julia D’SOUZA, John JACOB, Naomi S. SODERSTROM, “Nuclear decommissioning costs: The impact of recoverability risk on valuation”, Journal of Accounting and Economics, v.29, p.207-230 (2000). 12. Julia D’SOUZA, John JACOB, Naomi S. SODERSTROM, “Trident Utility: accounting for nuclear decommissioning costs”, Journal of Accounting Education, ed.18, 157-169 (2000). 13. Alan BOND, Juan PALERM, Paul HAIGH, “Public participation in EIA of nuclear power plant decommissioning projects: a case study analysis”, Environmental Impact Assessment Review, v.24, p.617-641 (2004). 14. Robert G. COCHRAN, Nicholas Tsoulfanidis COCHRAN, The Nuclear Fuel Cycle: Analysis and Management, 2nd ed. American Nuclear Society, La Grange Park, Illinois USA (1999). 15. P. BEZAK, V. DANISKA, I. REHAK, V. NECAS, “Algorithmization of nuclear installations equipment dismantling”, Nuclear Engineering and Design, v.240, p.4103-4110 (2010). 16. Sung-Kyun KIM, Hee-Sung PARK, Kune-Woo LEE, Chong-Hun JUNG, “Development of a digital mock-up system for selecting a decommissioning scenario”, Annals of Nuclear Energy, v.33, p.1227-1235 (2006). 17. Kwan-Seong JEONG, Kune-Woo LEE, Hyeon-Kyo LIM, “Risk assessment on hazards for decommissioning safety of a nuclear facility”, Annals of Nuclear Energy, v.37, p.17511762 (2010). 18. KwanSeong JEONG, DongGyu LEE, KuneWoo LEE, HyeonKyo LIM, “A qualitative identification and analisys of hazards, risks and operating procedures for a decommissioning safety assessment of a nuclear research reactor”, Annals of Nuclear Energy, v.35, p.19541962 (2008). 19. IAEA, Classification of Radioactive Waste, General Safety Guide, N° GSG-1, IAEA (2009). 20. M. I. OJOVAN, W. E. LEE, An Introduction to nuclear waste immobilisation, 2nd edition, Elsevier (2014). 21. Peter TATRANSKÝ, Vladimír NEČAS, “Conditional release of materials from decommissioning process into the environment in the form of steel railway tracks”, Nuclear Engineering and Design, v.239, p. 1155-1161 (2009). 22. T. HRNCIR, V. NECAS, “Recycling and reuse of very low level radioactive steel in motorway tunnel scenario”, Nuclear Engineering and Design, v.265, p.534-541 (2013). 23. OECD/NEA, International Structure for Decommissioning Costing (ISDC) of Nuclear Installations, OECD/NEA (2012).
INAC 2015, São Paulo, SP, Brazil.
24. M.C. BIERSCHBACH, “Estimating Pressurized Water Reactor Decommissioning Costs, A User’s Manual for the PWR Cost Estimating Computer Program (CECP) Software, Draft Report for Comment, Division of Regulatory Applications”, U.S.NRC, NUREG/CR-6054 PNL-8497, NRC (1993). 25. Jason A. GASTELUM, Steven SHORT, “Changes in Decommissioning Waste Disposal Costs at Low-Level Waste Burial Facilities”, U.S.NRC, NUREG/CR-1307, Rev.15, NRC (2012). 26. UFABC, Estimativa do custo mínimo de descomissionamento das usinas de Angra 1, 2 e 3, Centro de Engenharia, Modelagem e Ciências Sociais Aplicadas – CECS Programa de Pós-graduação em Energia e Engenharia de Energia, Universidade Federal do ABC, UFABC-DESCOM-EST-002-01, UFABC (2014b). 27. G VALREY, C. RUSCH, R2/R0-WTR Decommissioning Cost Comparison and Benchmarking Analysis, SKI Report, SKI (2001). 28. PGE-1061, "Trojan Nuclear Plant Decommissioning Plan and License Termination Plan (PGE-1078)", Revision 9, 2001. 29. MME/EPE, Plano Nacional de Energia 2030, Ministério de Minas e Energia, Secretaria de Planejamento e Desenvolvimento Energético, Empresa de Pesquisa Energética, PNE 2030, Novembro 2007; Brasília: MME: EPE (2007). 30. MME/EPE, Plano Nacional de Energia 2030, Geração Termonuclear, Ministério de Minas e Energia, Secretaria de Planejamento e Desenvolvimento Energético, Empresa de Pesquisa Energética, PNE 2030, Novembro 2007; Brasília: MME: EPE (2007).
INAC 2015, São Paulo, SP, Brazil.
