Version classiqueVersion mobile
OpenEdition Books

Proceedings of the fourth Resilience Engineering Symposium

 | 
Erik Hollnagel
, 
Éric Rigaud
, 
Denis Besnard

Managing Safety In Subcontractor Networks: The Case Of Olkiluoto3 Nuclear Power Plant Construction Project

Pia Oedewald, N. Gotcheva, T. Reiman, E. Pietikäinen et Luigi Macchi

Résumé

Subcontracting, especially large multicultural subcontractor networks challenge the traditional safety management practices. A resilient subcontractor network is characterized by its ability to provide and maintain an acceptable level of service in the face of operational challenges and disruptive events. The safety management in the nuclear industry is based largely on a traditional quality control paradigm which emphasized detailed planning, procedural adherence, surveillance of contractors and component based quality control. This paradigm has proven challenging in the Olkiluoto 3 nuclear power plant construction project in Finland and thus the case is a good example to discuss the practical needs and focus areas of engineering resilience in systems. We argue that the network’s abilities to respond, monitor, anticipate and learn are based on thorough understanding of the hazards related to the activities, connection of one’s work to safety and the systemic nature of safety. The conclusion opens new avenues for safety research by underlining the needs to; first, broaden the unit of analysis to inter-organizational level to explain and to dissolve the constraints of resilient activities and second, to develop practical means to analyze and develop the worker’s understanding of the hazards related to the activities.

Texte intégral

1 Introduction

1Subcontractors are nowadays common on large projects across industries, including the nuclear domain. Due to the refurbishments and the new nuclear power plant construction projects e.g. in Finland, the nuclear industry has faced a new challenge: to prove the safety and reliability of large, complex projects with a huge subcontractor network. As a prevalent business practice subcontracting offers certain benefits to the principal contractor, e.g. flexibility and cost efficiency (Wilpert, 2009). In addition to the benefits, however, it constitutes new challenges for organizational management in general and especially for safety management. In the nuclear industry the projects involve strict quality and safety requirements. Heterogeneous mosaic of subcontractors takes part in building and installing different structures, systems and components at a nuclear construction site. Many work tasks carried out by subcontractors during the construction stage could have important consequences for the reliability and safety of the operating plant.

2During the course of the Olkiluoto 3 project subcontractors have faced unexpected conditions and have found it difficult to proceed in their tasks exactly as planned. Thus, they have adjusted their activities based on their previous experiences from other industries (STUK, 2006). In some cases these adjustments have compromised the reliability of the technical structure or component they are working with since they didn’t take all the nuclear industry specific quality requirements into account. These kinds of adjustments have been interpreted e.g. by the media as poor safety culture and more surveillance has been required. We argue that in large complex systems the control and surveillance strategy needs to be at least complemented with approaches which help building resilience capabilities in the system. Based on the analysis of the challenges in the Olkiluoto 3 project we state that one important and underestimated factor in complex system’s resiliency is the ability of different stakeholders to understand the hazards, the functionalities and interconnections within the system. A practical resilience engineering question then is how could this understanding be identified and enhanced in organizations and even in large multicultural subcontractor networks?

2 Research Context

3Finland is currently constructing its fifth nuclear reactor Olkiluoto 3 (OL3) after thirty years since the previous reactor was commissioned. The Western Countries have little experience on nuclear new builds from the past decades since the Chernobyl accident in 1986 practically stopped the nuclear power programs in Europe and USA. The resent interest in new nuclear plants and major modernization of existing plants makes the industry attractive for international subcontractor companies. Insufficient awareness of the principles of achieving nuclear safety is a common issue among them though.

4The OL3 project is a turn-key delivery by an international AREVA-Siemens consortium. All and all the construction, manufacturing and installation stages require around 2000 subcontractor companies to provide the necessary components and services. The construction site at OL3 has workers from multiple different countries and it is estimated that they speak around 20 different languages.

5The OL3 power plant is purchased by a Finnish power company TVO which run two well performing nuclear power plants (OL1 and OL2). TVO holds the license to build the OL3 and the responsibility of its nuclear safety even though it is a turn key delivery. The regulator sets requirements for the licensee and controls the fulfilment of the requirements. It has multiple means for doing this, for example, reviewing and approving the design documentation of the safety critical systems, regular inspections on the licensee’s practices and, special inspections, participation in audits done at the subcontractors and event investigations. The Council of State requires all the organizations participating in design, construction and operating of the nuclear power plants to establish a good safety culture.

6The OL3 project is currently delayed for more than three years and an arbitration court is in process to define the financial responsibilities for the delay.

3 Method

7The analysis presented in this paper is based on the two authors (Oedewald & Reiman) experience in an event investigation, a case study and regulatory special investigations at OL3 construction site as well as on other studies and presentations about the OL3 project. In 2006 the Finnish Radiation and Nuclear Safety Authority (STUK) launched an investigation of management of safety requirements in the subcontractor network due to significant quality deficiencies in practices how the concreting of the reactor base slab was executed. Around half years after the concreting of the reactor base slab the licensee informed the regulator that the concrete mixture used was not according to the specifications since it was modified ad hoc during the pouring. The investigation team investigated the concreting case and two other cases where the communication and fulfillment of the safety requirements in the licensee-vendor-subcontractors chain had not been successful (STUK 2006, see also Koivula, 2008).

8In 2007 similar challenges emerged during the OL3 reactor containment steel liner welding. The subcontractors involved faced challenges in fulfilling the quality requirements of the weld and thus they ended up using improper repair methods. Due to the recurrent challenges at OL3 project we carried out a case study to investigate the OL3 licensee’s and the vendor’s communication to the subcontractors on safety and quality requirements (Oedewald et al. 2009). The case study was done by interviews and document analysis and the results were integrated with current literature of network management and safety culture. The main research questions were what is the role of the subcontractors’ knowledge and understanding in safety culture? How do the licensee and the vendor communicate the quality and safety requirements to the subcontractor chain? Could subcontractors’ knowledge and understanding of the functionalities of the work scope and the risks help the project to avoid quality and safety deficiencies? (Oedewald et al., 2009.)

9The researchers also participated as experts in three safety culture inspections at the OL3 construction site during 2008-2010. During the inspections the researchers interviewed field workers and their managers and safety culture experts. Further observations on working conditions and practices and documentation were done.

4 Results

10The investigation at OL3 in 2006 revealed that the network of actors did not share an understanding on how the construction phase can affect nuclear safety (STUK, 2006). Thus the requirements for safety culture, documentation and handling of deviations did not make complete sense to all parties. From their point of view OL3 site was just a construction site where they did their normal construction jobs. In fact most subcontractors believed that the required safety trainings and safety culture slogans meant occupational safety issues. Also the interviewees in managerial position had the conception that subcontractors, especially the workers (welders, fitters etc.) don’t need to know too much about nuclear safety e.g. how the component or structure they are working with relates to nuclear safety. It was emphasized that subcontractors are required to follow the plans and instructions. Further the managers emphasized that since the subcontractors are carefully selected they are professionals in their own tasks. However, only a couple of months after the construction work had started the quality audits and quality controls had detected hundreds of deviations (non conformances) and works started to be delayed (STUK, 2006). More quality control personnel were hired to intensify the quality control process but the more quality control work was done the more deviations were detected. Since the tasks of different subcontractors were interdependent a deviation in one stage sometimes caused widespread effects to other tasks. The overall project failed to follow the schedule from the beginning and multiple small deviations affected daily tasks in all levels of the network.

11It was clear that multiple adjustments needed to be made to plans and working practices due to these changes. However, while trying to adapt to the new situations the subcontractors sometimes failed to do it in a way which is acceptable from nuclear safety point of view. The regulator paid attention to the fact that the subcontractors did not always themselves seem to detect when their activities fulfill the specific requirements from quality and safety point of view. For that reason we carried out a study focusing on identifying the mechanisms with which the vendor and the licensee communicated the safety and quality requirements of the work to the contractors at the construction site (Oedewald et al., 2009).

12The study showed that there existed lots of documentation and multiple channels and for describing the quality requirements (i.e. materials, methods, different parameters, tolerances etc.) to the parties in the supply chain. However, the forums for discussing the rationale behind the quality requirements especially with the shop floor workers were insufficient. The documentation used by the subcontractors provided only a limited description of the functionality of a certain component and its link to the overall functioning of the power plant and thus nuclear safety. In other words the subcontractors receive documentation which describe quality requirements that may deviate from what they have used to seen in other constructions sites but the documentation doesn’t provide clear explanations what is the rationale of these special quality requirements. However, we found out that majority of the communication of the quality and safety issues took place in face to face situations, in kick off meetings, site meetings and tailored training for specific work scopes. This was not as effective as one could hope for due to e.g. language barriers, hierarchical cultures where only high level manager participates in the meetings and high turnover of the subcontractor’s staff (some meetings or training could be once a month but new workers arrive in the meanwhile).

13The study revealed that some subcontractor foremen were not aware of the safety functions of their work scope. They were not able to explain what could be the safety effects if they were not able to meet the quality requirements. In general they thought that safety is important at a nuclear construction site but not all of them could describe their own role for nuclear safety. They were not able to answer to all questions from their subordinates. Neither were they able to judge whether the adjustment they did concerning working methods or order of tasks could have some nuclear safety implications. They emphasized that it is an agreed practice to always inform their managers if they detect anything which may be a safety issue or something which could requires changes in the course of the work. One could raise a question however; how do the subcontractors identify a safety issue if they are not aware of what type of a system they are working with? Further, we found out that sometimes supervisors seemed to have challenges in motivating their subordinates to follow the rather complicated procedures because they were not able to understand the rationale behind them.

14In 2008 a former employee of one of the companies working at the OL3 construction site reported publicly his worries on the safety culture and quality of welding works. This initiated a series of safety culture inspections and resulted in clear demands from the regulator that the licensee needs to show that it monitors the safety culture of the OL3 construction site in a systematic way even though the vendor is managing the activities in the subcontractor network.

15The regulator took our case study results into account and explained that safety culture is not only positive values and attitudes towards safety but it requires understanding of the risks, interconnections and the safety effects of the activities. The questions around the sufficient level of understanding, e.g. exactly what each worker should know about the component and its effect on nuclear safety, have not been sufficiently resolved in OL3 project nor in the safety science in general. However, the challenges in OL3 initiated discussions in the nuclear community on the fact that in complex systems the workers role is not just to follow the plans and the instructions. The activities will not get done if that’s the only strategy. Further, the possible safety challenges or safety improvement initiatives may be missed if the workers are not given the opportunity to expand their focus and understand the effects of their work.

16The results of the OL3 investigations and case study pinpointed certain development needs in most safety culture models and helped us crystallizing our safety culture theory (see e.g. Reiman & Oedewald, 2009; Oedewald et al. in press). We state that safety culture is organization’s potential for safe activities. In addition to classical safety culture attributes such as values and mindset we emphasize that a good safety potential requires thorough understanding of the hazards and a systemic safety conception. No matter how good the intentions are and how much the organizations value safety they may end up in unwanted consequences if they don’t understand the hazards and the interconnections. Thirdly, well designed organizational structures and processes are needed in order maintain high quality of work in these dynamic and complex environments.

5 Discussion and Conclusion

17Despite rather clear safety management requirements and strong tradition in that domain, the nuclear community in Finland was not adequately prepared to the challenges of managing such a large network of heterogeneous actors from the beginning of the project. One contributing factor was that traditional safety management approach considers single organizations as a unit of analysis. The risk assessment practitioners define their scope even in a more narrow sense. When the activity is carried out by a multicultural network of companies new vulnerabilities arise from the complexity of communication, interdependencies between the actors and conflicting goals. On the one hand, the diversity in cultural background, knowledge and experience among subcontractor companies can be used to nurture network’s adaptive capacity. On the other hand, according to the principle of homophily in social networks, dissimilarities do not breed connections. Ruuska et al. (2009) state that large projects are characterized by (a) multiple organizations which are seeking success with different objectives, (b) changing priorities of project objectives, and (c) influence of wider sociopolitical environment.

18The nuclear safety focus was not evident to the OL3 project network. Most subcontractors involved in had no background in the nuclear industry. They were not familiar with the industry specific regulations nor did all of them understand the hazards and accident mechanisms of nuclear power production. The nuclear industry actors were blind to the culture outside their own industry. In the OL3 case the key organizations did have safety management systems but they didn’t help in creating shared understanding how to create shared understanding of the main goals or how to tackle the complexity and adjust activities in a reliable way. The multiple tight connections between various actors and activities at OL3 resulted in a “going solid” condition at times (Cook & Rasmussen 2005); activities in one area become significantly dependent on seemingly insignificant events in seemingly distant areas. Safety management emphasized the responsibility of an organization and the concept of an organization was understood as referring to the juridical entity, the company. If there are multiple companies involved in a shared project they form a virtual organization. They should align their practices to reduce the complexity and to boost development of knowledge and understanding among the actors.

19We state that the challenges experienced in the OL3 project can not be attributed to any single organization. Rather, the challenge was that the unit of analysis and management was too narrow and the generic lessons learned developed slowly. Network’s resiliency is captured in its ability to understand and absorb complexity, and to emerge from challenges with a greater repertoire of actions to draw from. Thus we propose that a resilient subcontractor network is characterized by its ability to provide and maintain an acceptable level of service in the face of operational challenges and disruptive events. The existing literature indicates that managing safety in subcontractor networks in high-hazard industries often remains problematic. Some authors point to the inadequate training, communications breakdowns, low levels of trust, and insufficient job commitment as challenging aspects of using subcontractors (Kochan et al., 1994), while others stress that higher risks exist due to the very nature of contract work, e.g. these workers are often undertrained and have limited competence in the specifics of a certain industry (Rousseau & Libuser, 1997). The resilience engineering literature often highlights the role of local initiative and adjustment as a source of resilience (e.g. Carvalho et al., 2006). Local adjustments took place at OL3 work. The OL3 case shows that the local adjustments need to be based on sufficient understanding of the system characteristics in order to support system safety. The role of experience in resilience has been emphasized in other case studies as well (see e.g. Cardiff et al. 2008) but the content of the knowledge the experience brings has seldom been described.

20When compared with the four abilities of resilient systems as defined by (Hollnagel, 2009) the OL3 project (including its stakeholders) faced challenges in all of those in the beginning of the project. They were not able to anticipate e.g. efforts needed to design certain stages with rather inexperienced parties. Further, the communication challenges related to the language barriers were not foreseen until they realized. The ability to respond to the topical issues was not optimal since the communication of challenges faced by subcontractors was mediated through the vendor to the licensee. The communication happened in some cases slowly and not in a completely open manner. The regulator stopped the work couple of times on site and required clarifications and improvements to ensure the quality of the activities. The ability to learn from experiences in such a huge project is challenging due to the number of activities going on. Very early in the project the quality database was overloaded with more or less significant findings. The project didn’t have a function which would have analyzed or trended the findings. Problems were tackled one by one. Only later the recurrent deviations and delays motivated the project organization to pay more attention to the generic topics around subcontractor management. This also resulted in developing more sophisticated means for monitoring e.g. the working culture at the construction site.

21The importance of ensuring safety culture among subcontractors already at the design and construction phase of NPP has been highlighted both by international authorities (e.g. IAEA, 2007) and the Nordic nuclear safety regulators. The challenges identified in the OL3 construction project have revealed that the nuclear community is still focused on technical reliability and not paying enough attention the social phenomena such as project management and safety culture (Koivula, 2008). When dealing with subcontractors, the traditional management approach is to control and closely supervise their work. However, our case study provides rich evidence that sticking only to such an approach could lead to problems, e.g. considerable delay in the planned time schedule.

22The need for development of a safety mindset and safe working practices within the project network should be emphasized (Reiman et al., 2011). Development of safety mindset cannot be forced with hierarchical orders but rather it requires intrinsic motivation. We state that knowledge and understanding on how each of the specific tasks can contribute to nuclear safety facilitate this motivation. A motivation to consider safety in all daily activities and in strategic decision making as well stems from understanding of what could go wrong.

23Understanding of safety as a complex and dynamic phenomenon could contribute to improving the resilience of the entire subcontractor network as resilience arise from the process of knowledge-sharing among different actors in the network. To foster this process, we propose that the unit of analysis should be changed in safety management from one organization to the actual set of organizations responsible for the activities. This also calls for designing new ways for cooperation to enhance mutual understanding among diverse subcontractors’ groups.

Bibliographie

References

Cardiff, K., Sheps, S., Nyce, J. & Dekker, S. (2008). Is safety quality? Is quality safety? Paper presented in 3rd Symposium on Resilience Engineering. Juan-les Pins, France, October 28-30 2008.

Carvalho, P., dos Santos, I., Gomes, J, da Silva Borges, R. & Huber, G. (2006). The Role of Nuclear Power Plant Operators’ Communications in Providing Resilience and Stability in System Operation. In proceedings of 2nd Symposium on Resilience Engineering. Juan-les-Pins, France, November 8-10 2006.

Cook, R. & Rasmussen, J. (2005). ‘‘Going solid’’: a model of system dynamics and consequences for patient safety. Quality and safety in health care, 14, 130-134.

Hollnagel, E. (2009) The four cornerstones of Resilience Engineering, in C.P. Nemeth, E. Hollnagel, & S. Dekker (Eds.) Resilience Engineering Perspectives, Vol.2, Preparation and Restoration, Surrey, UK, Ashgate Publishing.

Kochan, T., Smith, M., Wells, J, & Rebitzer, J. (1994). Human resource strategies and contingent workers: the case of safety and health in the petrochemical industry. Human Resource Management, 33(1), 55-77.

Koivula, N. (2008). Promotion of safety culture. Licensing and regulatory oversight of new nuclear build, Presentation workshop in Helsinki and Olkiluoto, 1-4 September, STUK report.

Oedewald, P., Pietikäinen, E. & Reiman, T. (in press). A guidebook for evaluating organizations in the nuclear industry -an example of safety culture evaluation. SSM research report series.

Oedewald, P., Reiman, T. & Talja, H. (2009) Safety culture: Understanding the safety significance of the job in the construction of the OL3 nuclear power plant, VTT research report.

Reiman, T. & Oedewald, P. (2009). Evaluating safety critical organizations. Focus on the nuclear industry. Swedish Radiation Safety Authority, Research Report 2009:12.

Reiman, T., Kahlbom, U., Pietikäinen, E. & Rollenhagen, C. (2011). Nuclear Safety Culture in Finland and Sweden – Developments and Challenges. NKS-239. Nordic nuclear safety research NKS, Roskilde, Denmark.

Rousseau, D.M. & Libuser, C. (1997). Contingent work in high risk environments, California Management Review, 39(2), 103-23.

Ruuska, I., Artto, K., Aaltonen, K. & Lehtonen, P. (2009). Dimensions of distance in a project network: Exploring Olkiluoto 3 nuclear power plant project, International Journal of Project Management 27: 142-153.

STUK (2006). Management of safety requirements in subcontracting during the olkiluoto 3 nuclear power plant construction phase. Investigation report 1/06. Radiation and Nuclear Safety Authority.

Wilpert, B. (2009). Impact of globalization on human work, Safety Science 47, 727-732.

Auteurs

VTT Technical Research Centre of Finland P.O. Box 1000 FI-02044, Finland E-mail: pia.oedewald@vtt.fi

VTT Technical Research Centre of Finland P.O. Box 1000 FI-02044, Finland E-mail: pia.oedewald@vtt.fi

VTT Technical Research Centre of Finland P.O. Box 1000 FI-02044, Finland E-mail: pia.oedewald@vtt.fi

VTT Technical Research Centre of Finland P.O. Box 1000 FI-02044, Finland E-mail: pia.oedewald@vtt.fi

VTT Technical Research Centre of Finland P.O. Box 1000 FI-02044, Finland E-mail: pia.oedewald@vtt.fi

© Presses des Mines, 2011

Conditions d’utilisation : http://www.openedition.org/6540