By Jukka Laaksonen, Director General, STUK – Radiation and Nuclear Safety Authority
Today Finland has 125 years of nuclear power reactor operating experience and this gives a good basis for construction and licensing of new nuclear power plants (NPPs). Two VVER-440 type plants are located in Loviisa and two BWR’s of Swedish (Asea-Atom) design are located in Olkiluoto. The initial net power of those power plants was 440 MW and 660 MW, respectively. The power of Loviisa reactors has been increased in one step and Olkiluoto reactors in two steps. Information on the operating nuclear power plants is provided in Table 1.
Licensing of a nuclear facility in Finland has three steps. The first step called Decision in Principle is a political one: the Government has to decide that a new nuclear power plant is in line with the overall good of the society and the Parliament has to give its consent to the decision. The Decision in Principle on a light water reactor of nor larger than 1,600 MW and located in Olkiluoto was approved in the Parliament in May 2002. Before the decision the utility TVO had completed the Environmental Impact Assessment. STUK had conducted a screening type safety assessment of several NPP designs to conclude that alternatives are available for constructing a NPP that could meet the Finnish safety requirements. The safety assessment process by STUK included discussions with interested vendors and resulted in design modifications that were proposed in the bidding phase.
The bidding process was started after the Decision in Principle and the main contract on Olkiluoto 3 was signed in December 2003. The site preparations started soon after that, and the construction permit for Olkiluoto 3 was issued in February 2005.
Olkiluoto 3 is the first EPR being constructed and it is a turnkey project offered by a consortium Areva-Siemens. Areva is responsible for the Reactor Island and Siemens for the Turbine Island.
The start of the project was very slow because the vendor was actually not adequately prepared when the construction permit was given. Parts of the detailed design and the working documents were not yet available and this hampered the planned progress in construction. It took also some time to find designers and experienced organizations for construction and manufacturing.
New advanced design features of the EPR, the increased unit size and application of new technologies brought major challenges for implementation. Many design features were implemented for the first time and some of the manufacturing and other technologies had not been proven in tests or in practical applications.
It took some time to organize the construction of the Reactor Island but after an experienced company had been contracted and the working design was available the civil construction started to proceed in a controlled manner. That happened about two years after the construction permit had been received. It remains to be seen whether the installations at the Reactor Island and the plant commissioning succeed with less initial inertia. In total, the schedule of the Reactor Island has in five years after the construction start slipped for about three years. The Turbine Island implementation has been more “business as usual” and has succeeded better.
In spite of the initial difficulties, the Olkiluoto 3 project has progressed reasonably well, and the negative impression created by the media reporting, often based on misunderstanding and repeated blame by those who want to prevent new nuclear build worldwide, is not well founded. Olkiluoto 3 is not a warning example but has provided valuable lessons for subsequent NPP construction in Europe and in the U.S.
Preparedness of the involved organizations for the new build
The staff of Nuclear Island vendor Areva had been strongly reduced from the time of earlier construction. Especially the number of designers was too small for quick start. On the other hand Areva had large economic resources for fast growth and it was expected to have a capability for relatively quick restart of nuclear build.
Vendor (Areva): Areva could offer a new design that was based on extensive operating experience in France and Germany. Development of the design had taken almost 10 years. Design work had been supported with results of experimental research and also by active involvement of the French and German regulatory bodies and their Technical Support Organizations. However, the design was still in a conceptual stage and the parties involved did not recognize how much additional work was needed to complete it.
A major drawback, although not clearly recognized at the time of signing the contract, was that Areva did not get full benefit of its earlier experience as a NPP supplier. Areva signed a turn-key contract on Olkiluoto 3 and took responsibility on Reactor Island without experienced partners and without having all needed competences. In all of Areva’s earlier projects, the owner and licensee of the French nuclear power plants EdF had played a key role of the architect-engineer and had been responsible for the construction management.
Many of the experienced nuclear manufacturers that had contributed to the earlier Areva projects had left the business. It was necessary to find new subcontractors and to coach them in the nuclear manufacturing.
Licensee (TVO): TVO has long-time experience from NPP operation and has implemented several modernization projects over the years. These include also up-rating the power of its plants. Throughout the 1980’s and 1990’s, TVO co-operated actively with the ABB Atom for developing an advanced new BWR design. It also conducted feasibility studies with other vendors and made an application to start new construction in the early 1990’s. This application was rejected on political grounds by the Parliament.
TVO’s call for tender got major benefit from the European Utility Requirements (EUR) that had been developed in more than 10 years’ time by the leading nuclear utilities in Europe. About 85percent of the technical requirements were taken directly from the EUR.
TVO’s key persons had worked in expert duties during construction and commissioning of the operating units but none of them had hands-on experience from management of a large construction project. It is evident that before signing the main contract, none of the two parties, Areva and TVO, adequately appreciated the key role of an experienced construction company for the success of the project. Furthermore, it seems that TVO was not adequately aware of the limitations in the capabilities of the potential vendors and the actual status of the available designs. Target set for the construction time in the call for bids was therefore not realistic.
Regulatory body (STUK): STUK had staff of adequate size and experience for making the construction permit review. Key persons had regulated already the construction of the operating NPPs.
STUK could rely on its own in-house competence for making safety assessment: reviewing safety analysis and conducting evaluation of the design and the management systems of involved parties. It had also staff for conducting inspection of structures and components. Furthermore, STUK had arrangements to request expert support from organizations that are able to conduct testing and independent analysis (deterministic, PRA).
STUK had been developing state-of-the-art national safety regulations since 1970´s, based on information received from extensive international co-operation. It had reviewed and assessed for more than 20 years the plans for modernization of operating plants, and also several feasibility studies intended to start new construction. In such reviews it had got extensive knowledge on different NPP designs.
A difficulty for a Areva was STUK’s unique regulatory approach with an early focus on quality of structures and components. This approach had been learned by vendors from the Soviet Union and Sweden in the earlier NPP construction projects in Finland but Areva was not used to such comprehensive assessment and inspection practice.
The original target was to connect the plant to the grid about four years after getting the construction permit and then complete the commissioning program for final turn-over to the licensee about 4 to 5 months later. Today, the target is to start commercial operation at the end of June 2012. Nuclear Island is thus about three years behind the original schedule.
Main reasons for the delay are
- Too ambitious original schedule for a plant that is first of its kind and larger than any NPP Ibuilt earlier,
- Inadequate completion of design and engineering work prior to start of construction,
- Shortage of experienced designers,
- Lack of experience of parties in managing a large construction project, and
- Worldwide shortage of qualified equipment manufacturers.
Construction of the Turbine Island that is under responsibility of Siemens has progressed much better. From the very beginning there has been close cooperation between Siemens and its subcontractor, an experienced construction company. This has resulted in good integration of design and construction work. Installations of the main equipment at the Turbine Island were completed about one year behind original schedule and the equipment are now in conserved state waiting for commissioning to start.
In planning and scheduling new build, it is necessary to recognize that circumstances in Europe and the U.S. are quite different from the 1970’s when most of the currently operating plants were constructed.
Vendors of the 1970´s had large experienced organizations and these incorporated comprehensive in-house capability for design and manufacturing. This reduced the dependence on subcontractors, and management of the projects was more straightforward.
In the 1970’s, there was enough of skilled manufacturing capacity in the market, and the large volume of construction facilitated finding experienced project managers. What’s more, need for design effort and for qualification of design features was smaller because the designs were often based on work done earlier in similar projects.
Since the first era of intensive NPP construction, vendors have lost much knowledge and skills when experienced experts have retired. Today, new types of competencies are needed for new technologies such as digital instrumentation and control systems. A good company name earned in the past is no guarantee for success. More important is the experience and competence of individuals actually assigned to the project.
For starting new build, vendors need to establish a sub-contractor network from companies with proven skills. Awareness of nuclear quality and understanding of nuclear safety culture must be taught to organizations that have no previous nuclear experience. Management of work conducted by the sub-contractors is a challenge of its own.
The situation in Japan and Korea is evidently quite different; vendors have been constructing without a break and have had manufacturing of the main equipment under their direct control. Early contacts between vendors, licensee and regulator were of great benefit Feasibility studies of several designs in the early stage of the project were found very useful and facilitated subsequent licensing process. These studies provided
- Technical discussions between potential vendors, license applicant and regulatory body
- Early allocation of licensee and regulatory resources to the safety assessment of each alternative design
- Identification of crucial safety issues before and during the Decision in Principle process
- Vendors addressing the safety issues that the regulatory body had raised, by improving the design or by providing additional safety evidence in parallel with the bidding and bid evaluation process.
Each design proposed in bidding was improved from the original version that had been presented for tentative review during the Decision in Principle process. Making safety requirements and the regulatory practice clearly understood to the vendor is needed to avoid uncertainties in licensing and regulatory oversight.
Making safety requirements and the regulatory practice clearly understood to the vendor is needed to avoid uncertainties in licensing and regulatory oversight
European Utility Requirements were used to present most of the technical requirements to potential bidders, but these did not include all necessary national safety requirements. The licensee and the regulator need to discuss early enough on how the national safety requirements can be best presented in the call for bids. It became evident during construction that just making reference to the national requirements and the regulatory guides is not adequate to ensure that requirements are correctly understood by vendors. Discussion in preparing for the Olkiluoto 3 project should have been more extensive to agree on how to express clearly the national safety and quality requirements and to describe the key features of the regulatory process in the call for bids.
For ensuring smooth progress of the project, all parties (vendor, licensee and regulatory body) should be familiar with the licensing, regulatory oversight and inspection practices both in the vendor country and in the customer country. The differences and their implications should be discussed between the parties for optimized regulation of the project. The vendor needs to understand and take seriously the national regulatory practice.
In Finland, the regulatory practice is different from what Areva had met elsewhere. For instance, a so-called Construction Plan including design and manufacturing information and QC plan must be approved by both the licensee and STUK for each non-standard equipment (for example, other than valves, pumps, electrical motors) in safety class 1 and 2, before manufacturing or construction is allowed to start, and inspections with hold points are made during construction and manufacturing.
Preparedness of parties for project implementation must be known before start
In order to avoid delays and difficulties in the project implementation, it is necessary to allocate enough time for planning stage and to assess the preparedness of each party before starting construction. Each of the parties (vendor, licensee and regulatory body) should conclude that licensee’s capabilities, arrangements for implementation and resources are adequate; vendor’s capabilities, arrangements for implementation and resources are adequate; design has been completed to a level required for a controlled construction start and for smooth implementation; qualified subcontractors are available as needed; and plans and contracts exist for managing the subcontractor chains.
Both the licensee and the vendor must have
- Project management and quality management skills
- Experience from management of a large construction project
- Knowledge and experience in all technical areas relevant for nuclear safety: civil, mechanical, electrical and I&C engineering
- Knowledge and experience on nuclear technologies (water chemistry, nuclear fuel, reactor physics, thermo-hydraulics and safety analysis)
- Skills and arrangements to verify achievement of required quality, and
- Arrangements to control and correct quality non-conformances.
In addition, the vendor must have at its disposal experienced designers who have a realistic view on the actual challenges involved in implementation. It also needs a confirmed access to manufacturers and constructors who have proven capability to meet the designer’s intent and the related specifications.
Experience and skills for construction management are necessary for project success
Slow start of Olkiluoto 3 construction indicated that the licensee and vendor did not recognize the importance of skills and experience to manage a major construction project. It was understood only after the initial poor performance in construction that although Nuclear Island vendor is experienced it has not taken direct responsibility for construction in its earlier projects.
For managing the construction, both parties need to know how to schedule the work, what resources are needed and when, how the vendor can find competent contractors and how it should manage them and how the licensee should conduct its oversight.
Design must be completed well before respective implementation
Inadequate completion of design and engineering work prior to start of construction is detrimental to implementation of the project in targeted schedule. It delays the start of construction activities at full speed and leads to attempts to reschedule manufacturing and construction steps. Such attempts make the project management complicated and all delays cause continuous time and cost pressures to the involved organizations.
In Olkiluoto 3 project, detailed design has been done too late. Delivery of Construction Plans to STUK’s review and approval has often been delayed from the planned schedule. Even worse is that fixing the sometimes-inadequate quality of design and engineering has delayed start of manufacturing and thus caused major difficulties for project management. Rotation of the initially poor Construction Plans for corrections and reassessment has been time and resource consuming. Insufficient Construction Plan quality has also caused numerous unnecessary comments and this has required extra time in the approval process. Several successive document revisions have made subsequent inspections at vendor’s premises complicated.
Conceptual design provides definition of safety design criteria and their application in safety systems redundancies, diversity, physical separation and protection from external hazards. Layout of buildings, process diagrams of main systems and key parameters for main equipment should be fixed.
Basic design provides deterministic and probabilistic safety analysis, loads for design of buildings and specification of system parameters and limits for protection system. Diagrams of main systems, drawings of main buildings and 3D-drawings of main fluid systems and structures should be presented.
Basic design needs to be completed before construction license application.
Detailed design provides design specifications of all components and diagrams of all systems, drawings of all buildings, including 3D-drawings of all fluid systems and structures. Detailed design needs to be conducted in parallel with the Construction License review and be completed not later than about two years after the construction start.
Working design provides shop drawings and plans for manufacturing of the equipment, material specifications and quality control plans. For buildings it provides architectural drawings and specifications for construction. It needs to be completed well before start of respective manufacturing or construction work where it is needed, and also permitting time for possibly required review by the licensee and regulatory body.
Vendor needs to have fair partnership with the sub-contractors
For ensuring good management of the subcontractor chains, it is important that in each call for tender on sub-contracts the vendor clearly indicates and emphasizes the nuclear specific practices and requirements. These could include:
If the nuclear specific practices are not recognized and understood by the sub-contractors at the time of signing the contract, difficulties are to be expected in a later stage.
Verification of the sub-contractor capabilities before making purchase is important
It has been noted in connection with many Olkiluoto 3 sub-contracts that the actual competence of manufacturers and sub-contractors is not easy to judge through auditing only. Therefore an evaluation of the manufacturers’ ability at the shop floors is important.
The licensee needs to have means to ascertain that the issues specific to nuclear safety and quality management, and the respective controls are properly agreed in each contract between the vendor and its sub-contractors.
Capabilities need to be assessed not only in case of sub-contractors that are newcomers to the nuclear field. In connection with the Olkiluoto 3 project it has been observed that a manufacturer having earlier experience from nuclear work has lost its actual control on quality management at the shop floor level. In that case the workers tried to eliminate minor defects from major components, without understanding the potential harmful consequences and did not ask for specialist advice before taking action. This led to a serious consideration on a possible need to remanufacture the impacted components. Remanufacturing would have caused an additional delay of more than a year in the Olkiluoto 3 start-up.
Lack of communication among designers can lead to a disaster
If design work is conducted by different organisations and in different places (or even in different countries), good coordination and communication among the designers is vital for a successful outcome. Lack of coordination and communication within the vendor consortium has been a problem area especially in the early stage of the Olkiluoto 3 project and throughout the I&C design process.
Licensee and regulator should audit and carefully assess the communication process and the adequacy of communication between those designers who are expected to interact to ensure a consistent design.
New advanced safety features are not easily implemented
Qualification of a new construction or manufacturing method may take time if it is not done before the project start. For instance, new welding solutions were a challenge during the reactor pressure vessel manufacturing. Full-sized welds were made and inspected in test pieces prior to the actual production of the new type bi-metal welds that join the stainless steel safe-ends (same material as the coolant lines) to the vessel nozzles, made of carbon steel. The nozzles themselves were fitted to the vessel in a new geometry that gives an opportunity for in-service-inspections. This required development of an innovative automatic welding tool that was not easy to operate. The first production welds had quite large defects and repair welding was needed. Fortunately, the welds produced after the first ones succeeded well when the process had reached maturity.
Major difficulties were met in manufacturing of the main coolant lines. The design target was to reduce the number of welds to a possible minimum, which is a clear improvement compared with the operating plant generation. Achievement of the specified material properties was difficult and several pieces had to be re-manufactured before succeeding. Another concern was met in some of the shop welds that still had to be made in the main coolant lines. Micro-cracking that had not been faced by the manufacturer before was seen in the surface of the weld. Fortunately it could be demonstrated that the indications were only in the surface. They were removed by grinding and re-welding.
Also, many other large components for Olkiluoto 3 had to be re-manufactured once or twice to achieve specified quality and to ensure 60 years lifetime. Among these were most of the pressurizer forgings, some reactor vessel internals, some main circulation pump casings and shafts, high pressure turbine shaft, and stator of the main generator.
Licensee is responsible for the safety of its plant when it starts to operate
In order to bear its ultimate responsibility for safety, the licensee must have strong control of the project during construction. Turn key and fixed price projects are no exception. Similar quality control and auditing arrangements are needed independently of the contract type. It is therefore necessary that the respective roles and responsibilities of the licensee and the vendor are specified for the construction phase.
Licensee should conduct its own safety assessment to verify that the plant and its systems, structures and components are licensable. For this it needs a requirement management system of its own and an independent capability to verify and prove that all requirements are met. The licensee must also have a system for reporting and resolving all non-conformances that are identified in the audits and quality inspections. The experience from Olkiluoto 3 is that one could expect thousands of smaller and bigger non-conformances. Therefore it is necessary to have unambiguous requirements for classification of non-conformances.
Licensee management needs to be committed to implement a strong quality system
The system for quality management needs to provide transparent links between safety and quality classes and emphasize the general requirement on application of respective nuclear specific quality standards.
The licensee management has to communicate its requirements and expectations with respect to quality to all parties. The experience from Olkiluoto 3 also indicates that it would be worthwhile for the licensee management to have an opportunity to require that the vendor uses proven state-of-the-art technology in manufacturing and construction and selects manufacturers with proven good performance. The licensee management should not be satisfied to simply accept low cost technologies and manufacturers that barely meet the minimum agreed quality requirements.
Safety culture needs to be developed already during construction time
Strong message and transparent actions and decisions are expected from the management of the vendor and the licensee to promote safety culture: “safety and quality have higher priority than costs and schedule”. This needs to be demonstrated, for instance, in:
- Choice of qualified subcontractors
- State-of-the-art tools and methods
- Uncompromising compliance with the agreed requirements, and
- Walk downs by the management.
A questioning attitude is needed on every level and organisation, including the licensee, vendor and all subcontractors.
Workers should be encouraged to report potential safety concerns to their supervisors, and all questions and concerns raised by them need to be responded properly, also in case that a concern is based on misunderstanding on the worker side. Each person attending the project needs to understand the safety significance of his/her work, to promote personal responsibility.
Close regulatory oversight has been found to promote quality of construction
Throughout the Olkiluoto 3 project there have been multiple quality controls, carried out by manufacturers themselves, Areva, an independent third party, TVO and STUK. Therefore the product deviations have generally been detected with high sensitivity.
Nevertheless, in some situations the QC inspectors by the manufacturer, vendor and licensee have evidently faced too much economic pressure and have not been strong enough to enforce stopping of work and making necessary timely corrections. This has happened even when the work has not progressed as expected and the parties have recognized the deviation from the specified target. In such situation an intervention by a regulatory inspector has been needed.
Stringent regulatory approach and inspections are thus needed to verify that new manufacturing techniques and new type of equipment meet the specifications set by the designer.
Starting new build is demanding because much of the earlier experience and resources have been lost from the nuclear industry. Adequate time has to be allocated to good preparation of the project before actual construction start:
- Making design as early as needed for smooth construction,
- Qualifying the new design features and technologies,
- Building competent organizations,
- Specifying responsibilities of parties,
- Ensuring availability of qualified designers, constructors and manufacturers to implement the project, and
- Resolving potential regulatory uncertainties.
During the construction of Olkiluoto 3, STUK has concluded that close monitoring and oversight by both the licensee and the regulatory body is necessary to ensure achievement of specified quality, that is, meeting the technical standards and criteria that the vendor has specified and that have been approved as part of licensing and design documents.
Encouraging progress has been made during the project, and after the “teething problems” the construction has proceed well. However, in the first piping installations at the Reactor Island in October 2009, enforcement actions were again needed to make the welders follow the specifications given in the procedures.
While there have been many non-conformances and re-manufacturing needs, the quality awareness and pro-activity of the licensee and the manufacturers have been at a reasonably good level. The corrective actions have been taken in line with the QA/QC practices specified for the project.
The final quality in Olkiluoto 3 structures and components has not been compromised although in some cases achieving and proving the expected quality has required special efforts. These have included:
The observed difficulties at the construction stage have not raised concerns on the safety of the power plant when it will be ready to operate.
Editor’s Note: This article is derived from a conference paper given by Jukka Laaksonen at Nuclear Power Europe in Amsterdam in June 2010.
Nuclear Power International Issue Archives
View Power Generation Articles on PennEnergy.com