Giving VVERs a new lifetime30 June 1998
In the original VVER design, lifetimes were specified separately for the entire plant, for the reactor pressure vessel, and for a number of other components. The Finnish view on VVER-440 ageing, however, is that the vendor’s guidance would be used only as the starting point for the utility’s own development of a plant life management strategy. The Loviisa VVER plant management concluded that it is necessary to acquire and maintain thorough technical knowledge of the plant and, in particular, its maintenance and operating history. The Russians are also renewing their approach.
VVER-440 plants were originally designed for 25 or 30 years lifetime (see following list of currently operating plants).
In Finland, the 30 years lifetime was understood from the very beginning to be a minimum time guaranteed by the vendor in the main contract. This contract also specified how many transients of each type were used as the design basis. Each transient was postulated to cause certain thermal and hydraulic loads to the main equipment, and a fatigue analysis was provided by the designer to demonstrate that all of the load cycles, when taken together, could be tolerated with an adequate safety margin during plant lifetime.
There was much conservatism in this approach: the number of most load cycles was postulated to be significantly higher than actual operating experience, and for some transients the fatigue analysis used overestimated loads. For example, the Loviisa main contract specified the following numbers of main transients (the numbers in parentheses show the actual number of recorded transients at Loviisa 1 during its first 20 years of operation):
Heating up from cold shutdown to power operation and cooling to cold shutdown 300 (48)
Reactor scrams at power 500 (13)
Step change from full power to house operation 90 (6)
Start up from hot standby after reactor scram 600 (13)
Pressure test 20 (5)
Leaktightness test 70 (50)
By-pass of high pressure pre-heater 1600 (104)
This list is intended only to give an idea of typical transient numbers and does not include all transients which were accounted for in the fatigue analysis. From the above transients, the only one which has occurred with a frequency close to the design value is the leaktightness test. However, its contribution to fatigue is minor when compared with other transients.
If the actual figure for any of the transients starts to approach the design basis figure, the validity of the design basis can be verified at any time, or even a new modified design basis can be defined, by making a complete fatigue analysis where the postulated transient numbers are closer to the actual operating history. Such an updated fatigue analysis will be done for Loviisa before the age of 30 years is reached.
Separately from the guaranteed plant lifetime of 30 years, the reactor pressure vessel safety was guaranteed for 40 years of operation. As a potential life limiting factor, the design analysis evaluated the rate of reactor vessel wall embrittlement due to exposure to high energy neutrons. The analysis was based on research results from the first half of the 1970s which indicated adequate toughness of the vessel wall material for the entire lifetime. However, the first tests of surveillance samples irradiated in the Loviisa 1 reactor vessel for one year, pointed to a faster than predicted embrittlement. Similar concerns were later raised for most of the VVER-440 reactor vessels, due to high copper and phosphorus impurities in the circular weld next to the core. The oldest vessels in particular were found to be unacceptable for 40 years’ operation, unless corrective actions were taken to reduce the risk of vessel brittle fracture. An efficient remedy for recovering the wall toughness is thermal annealing, which is much easier to do in VVER-440 vessels than in Western PWR vessels. A number of VVER-440 vessels have now been annealed, and a series of other measures reducing brittle fracture risk have been taken. After corrective actions, reactor pressure vessel embrittlement is no longer considered a concern which might require an early closure of a VVER-440 plant.
In addition to the design lifetime of the entire VVER-440 plant and its reactor pressure vessel, a Soviet national rule had required that safety relevant components be given component specific lifetimes (the current version of that rule is PN AE G-7-008-89). The lifetimes were specified in the “component passports” issued by the component manufacturers. After the time given in the passport had expired, the component had to be replaced with a new one. This approach emphasised the responsibility of the manufacturer for his product even during operation, unlike the Western approach where the holder of the operating licence has an unshared responsibility for his plant.
The way of defining component lifetimes in advance was typical of the entire maintenance approach in the Soviet Union: instead of condition monitoring and corresponding maintenance measures, each component had been given a fixed frequency for servicing and for complete overhauls, and a pre-determined time of replacement.
Examples of component lifetimes were:
Control rod drives 5 years
Control rod absorber parts 4 years
Diesel engines 25 years
Cables 25 years
Batteries 20 years
A general observation from the component lifetimes is that for the mechanical components the times were often quite short and thus conservative, while for the electrical equipment they were too long.
The extension of component lifetime beyond passport specification was in principle permitted at Russian plants if a consensus decision could be made together by the operating organisation, design organisation, component manufacturer, and an authorised material examination organisation.
However, the national rules gave no practical guidance on what was required to prolong the lifetime. The manufacturers were reluctant to agree on any extension because their main interest was to sell replacement components. A common practice was thus to replace a component after its design lifetime had expired, no matter how good it still was for further operation.
CHANGING METHODS IN RUSSIA
The restructuring of the Russian economy now requires commercial optimisation of activities, as in any market economy. Deviation from the old traditional working methods has therefore become necessary.
The new nuclear energy law emphasises the responsibility of the operating organisation and consequently gives it more freedom to make its own judgements. Manufacturer’s guarantee is no longer required as a prerequisite of component life extension. This makes it possible to postpone component replacement if monitoring confirms it to be in good condition.
A national level document on how to manage component lifetime prolongation projects has been issued by Rosenergoatom and VNIIAES. Rosenergoatom is also preparing technical guidance on this topic.
As concerns the lifetime of the entire plant, Rosenergoatom has already started a project aiming for life extension of its VVER-440 plants. It has been reported even in public that the objective is to operate until the year 2010 those VVER-440 plants which were due to shutdown between 2001 and 2004.
The programme for life extension includes both research and improvement of plant systems and components. For example, at the oldest Kola unit the entire reactor protection system, with all of its associated equipment, was replaced with a modern one during the second half of 1996, and also a large part of the main control room monitoring devices were replaced with an advanced system. This suggests that permanent shutdown in 2003 is not to be expected.
The Russian regulatory body GAN is preparing new rules for a life extension procedure. Its aim is to do a thorough licensing review before new licences are granted. GAN has also said that a significant safety upgrading is a prerequisite for life extension. Critical issues for licensing include:
• Reactor pressure vessel embrittlement.
• Modernisation and upgrading of safety systems.
• Improvement of seismic resistance.
According to GAN, modernisation and extensive equipment replacement is needed in several systems including electrical, instrumentation and control, fire protection, and ventilation.
Some of those systems are already too old and their faults disturb stable operation of the plants.
THE SITUATION AT LOVIISA
The “guaranteed” lifetimes of the VVER components and the entire plant had little practical meaning in Finland because the operating organisation had to take full responsibility for their plant and had, in any case, to establish means for condition monitoring. It is also our general view that life management is not so much connected with the plant type, but reflects more the traditions and culture of the utility. A typical feature of Finnish life management is that, in addition to normal maintenance, deteriorating or obsolete equipment are being replaced gradually during each refuelling outage. Typical annual investment on new equipment at Loviisa has been $10-20 million (for two units).
A nuclear plant operating licence is issued in Finland for a limited time and can be renewed only after a thorough safety review which includes an assessment of ageing. The current operating licence of both Loviisa units was issued in March 1998 and will expire at the end of 2007, almost 31 years since Loviisa 1 was first connected to the grid. In the application the licensee said that the current estimate for the total lifetime is at least 45 years. The possibility for further extension will be investigated when the 45 year milestone comes closer.
The situation in Finland, as concerns the anticipated lifetime of the plant and the factors having most influence on it, cannot be regarded as representative of VVER-440 plants in general. There are some initial differences in the main components and the layout, but even more significant is a different operation and maintenance philosophy which has been followed since initial start-up. Due to different operating practice we can expect that the ageing of Loviisa’s main equipment proceeds more slowly than at most other VVER-440 plants. Also, the selective continuous investment on new equipment can make a big difference.
The most critical components affecting plant lifetime in Loviisa are the steam generators which are very difficult to replace due to their location deep inside the containment. At other VVER-440 plants the layout is not a similar obstacle for steam generator replacement. Other VVER-440s seem to consider the reactor pressure vessel replacement almost impossible, but in Loviisa this option is not excluded. In fact, the plant has in its stores a complete reactor vessel fabricated by Skoda, although its use as a replacement part has not been discussed seriously. The vessel was originally intended for a plant under construction in Poland but never finished.
Another component which may become critical in the long term is the main coolant pump. The 12 pumps at the two Loviisa units, plus the spare pumps, are the only ones of their kind, and maintaining knowledge on the specific purpose and importance of each design detail is not in the interest of the original supplier. Therefore, the operating organisation concluded that it will acquire and maintain thorough technical knowledge on these and other mechanical components which no longer have a competent supplier in the market, and cannot be easily replaced with a component from a different manufacturer.
Regarding plant operation and maintenance philosophy, the attitude of the Loviisa plant management has been, from the time that the plant was under construction, that the vendor’s guidance on operation and maintenance would be used only as the starting point for developing its own approach. This position developed progressively from the time that the operating staff were in training at Novovoronesh, where they were astonished to see how harsh a manner Russian operators treated the plant. For example, larger than necessary thermal loads were placed on the primary circuit equipment during normal cooldown, and excessive force was often applied to bolts in flange joints when their leaktightness could not be achieved as planned.
When developing the lifetime management approach for Loviisa, it was emphasised that relevant operation and maintenance activities must be incorporated and not treated as separate issues. Support to lifetime management is provided by continuous improvement of inspection and monitoring methods and by an extensive research programme which studies the ageing of representative surveillance samples, such as pressure bearing materials, I&C parts, cables and concrete structures.
Regarding plant operation the management stresses the importance of gently treating all equipment during normal operation and under test conditions, ie limiting to the greatest extent possible the number and magnitude of thermal and hydraulic transient loads and maintaining the water chemistry parameters in all cooling systems within their specified limits.
Operating and test procedures have been developed actively by the entire operating organisation with the aim of reducing equipment loads. Advanced process computers were particularly useful in collecting information from the extensive array of instrumentation and presenting in graphic form the thermal and hydraulic load histories of the equipment in different operating situations. The most critical pipework, such as pressurised surge lines, are equipped with surface thermocouples. In addition, in many other piping sections of the auxiliary and support systems, the fluid temperatures are recorded over short intervals and can be shown for at least qualitative assessments of loads whenever needed.
In the water chemistry area, not only is the primary coolant under careful control but the secondary circuit water quality has also received great attention throughout the plant’s operating life. Today, all Loviisa steam generators are still in an excellent condition – only two tubes at the entire plant have required plugging. One tube was plugged due to a fabrication fault and the other one due to wall thinning (the reason for thinning is not known). The situation at other VVER-440 units is variable: there are plants whose steam generators are in excellent condition, but also some with many plugged tubes.
The maintenance strategy at Loviisa is based on:
• Assessing the equipment condition from reliability records and from results of inservice inspections, tests, and on-line monitoring.
• Actively pursuing bilateral exchange of information with other VVER-440 plants, and incorporatiing lessons learned from other plants into the test and maintenance programme.
• Redefining service and overhaul intervals as need arises from the condition assessment.
• Repairing or replacing equipment before failure.
Reliability and maintenance records have been used actively to support decision making. Based on the collected data, replacement of some equipment was decided early in the lifetime. As examples, limit switches of the reactor protection system were changed after the first operating year, while the emergency feedwater pumps, containment spray pumps and steam generator safety valves were changed in two steps after 4 and 6 operating years at Loviisa 1. It should be pointed out here that many items of Loviisa’s key safety equipment are different from a standard VVER-440 plant, eg the primary coolant pumps, pressuriser safety valves, most control valves, and all containment isolation valves including fast isolation valves in the steam lines and feedwater lines.
In order to emphasise life management as one of Loviisa’s key issues, a systematic management programme has been developed and documented in a plant procedure. Features of the programme include the following:
• Responsibility for the whole programme is assigned to the maintenance manager.
• The task of continuous system specific monitoring of ageing is assigned to a number of engineers who are “owners” of the respective systems.
• Following an annual evaluation and discussion of the ageing situation, an updating of the life management programme is decided at a meeting which is attended by experts in ageing issues and by managers responsible for operation, maintenance, technical support and training.
• Special task forces have been formed for primary and secondary circuit condition monitoring, consisting of experts from all relevant technical fields.
• The plant’s reporting and data file systems saves and processes all relevant data on individual components and can provide information in different forms to the users; these comprise a basic data file for plant lifetime management (eg description of potential failure and ageing mechanisms and testing methods, test history, operation histories), maintenance history files, in-service inspection files (ASME XI inspections, other inspections such as erosion–corrosion monitoring of secondary circuits), a company wide file on equipment failures (which also covers conventional power plants operated by the utility).
The following are examples of equipment replacement practices at Loviisa 1:
• Instead of replacement of complete control rod drives after five years as recommended by the manufacturer, a programme of periodic inspections and overhaul has been implemented which has identified certain parts which must be periodically changed. Some new control rod drive systems have been acquired and installed to permit overhaul of drives during power operation, but none of the original drives has yet been removed from service.
• A replacement programme of pressure and pressure difference transmitters was carried out in 1986-92, involving 96 transmitters.
• A programme to replace 136 instrument penetrations (ie containment wall penetrations) and 33 electrical penetrations was carried out in the period 1987-1992.
• A programme to replace or to reconstruct valves and/or their actuators was carried out between 1987 and 1996 involving about 300 valves. The main reason was not ageing but poor initial qualification of the components to potential accident conditions, and also mismatches in the choice of actuators to the valves (in most cases we were concerned that the particularly strong motors could have broken the valves in case of torque limitation failure).
• The replacement of diesel engines has not been considered, but complete overhauls have been made one-by-one off-site, starting after 20 years of operation.
• Batteries were replaced after 10 years because there were signs of decreasing reliability.
• The main process computers and control room monitors were replaced after 11 years of operation to benefit from developments in computer technology.
• Surveillance samples of cables used inside the containment have been kept in representative locations and some of them have been tested at five year intervals. Some cables were replaced after 20 years operation because there was indication of thermal ageing.
• Secondary circuit piping and heat exchangers have been replaced based on thinning measurements. It should be noted however that the Russian steel in the secondary circuit piping in general provides good resistance to erosioncorrosion because it is slightly alloyed with chromium. Of the two feedwater line ruptures which have occurred in Loviisa, the first one was a consequence of erosion in a small fitting made of Western unalloyed carbon steel, and the second was the result of an erroneous design change which caused a strong eroding vortex.