Performance of Ontario’s CANDU nuclear generating stations in 2015

By: Donald Jones, P.Eng., retired nuclear industry engineer, 2016 March 19

At the end of 2015 Darlington had a four unit average lifetime Capacity Factor (CF) of 83.6 percent and an average annual CF of 76.1 percent. Bruce A had a four unit average lifetime CF of 69 percent and an average annual CF of 86.1 percent. Bruce B had a four unit average lifetime CF of 83.5 percent and an average annual CF of 84.4 percent. The six unit Pickering station had a six unit average lifetime CF of 72.4 percent and an average annual CF of 78.6 percent. Performance data for 2014 are discussed in reference 1.

The raw performance data for 2015 are taken from the Power Reactor Information System (PRIS) database of the International Atomic Energy Agency (IAEA). Note that the Load Factor term used in the PRIS database has the same meaning as CF. CFs are based on the (net) Reference Unit Power and on the (net) Electricity Supplied, as defined in the PRIS database.

The performance of some of Ontario’s nuclear generating stations is affected by the surplus of generation in the province. The surplus usually arises because of unreliable intermittent wind generation coming in at times of low demand and wind generation is expected to increase even more over the next several years. Some nuclear units saw electricity output reductions during periods of surplus baseload generation (SBG). This means the CFs are not a true performance indicator for those units (reference 2). A better metric of performance in these cases would be the Unit Capability Factor (UCF – used by Ontario Power Generation and by Bruce Power) or the Energy Availability Factors (EAF) that are shown in the PRIS database. The EAF adjusts the available energy generation for energy losses attributed to plant management and for external energy losses beyond the control of plant management while the UCF only includes energy losses attributed to plant management and excludes the external losses attributed to grid related unavailability and other things. This means that on unreliable grids, for example, UCF will be significantly higher than EAF but for Ontario there will be no significant difference. The UCF and the EAF take into account reductions in plant output due to load cycling and load following. For units that load cycle and/or load follow the CF will be significantly lower than the EAF. For example, Bruce B unit 5 has a 2015 annual CF of 86.4 percent and an EAF of 91 percent. The UCF and the EAF are based on reference ambient conditions so, unlike the CF, they cannot exceed 100 percent. In some cases the CF is more than the EAF because of lower than design cooling water temperatures that increase the electrical output of the unit. The only reason for using the EAF here (see later for Bruce units) instead of the UCF is that EAFs are now available in PRIS and UCFs are not presently available (well, the author could not find them).

1,000 day breaker-to-breaker run is possible with CANDU

By: Donald Jones, P.Eng., retired nuclear industry engineer – 2015 June 22

How far can we extend the continuous on-line power operation (breaker-to-breaker runs) of the world’s commercial Generation II and Generation III nuclear power plants. Is 1,000 days possible?

To date the world record for PWR (Pressurized Water Reactor) continuous on-line operation is the 705 day run by Three Mile Island unit 1 an 819 MWe (net) unit in the U.S. that went into commercial operation in 1974 September. The run ended in 2009 October when the unit went into a planned refuelling outage. This run broke the previous world record of 692 days of another PWR, Calvert Cliffs unit 2, an 850 MWe (net) unit in the U.S. that was put into commercial operation in 1977 April. This run ended in 2009 February with a refuelling outage.

LaSalle unit 1, a 1137 MWe (net) unit in the U.S., that was put into commercial operation 1984 January, holds the world record for a BWR (Boiling Water Reactor) with 739 days when it came off-line in 2006 February. As it happens its twin, LaSalle unit 2, became the second place world record holder when it completed a run of 711 days on 2007 February. LaSalle unit 2 went into commercial operation in 1984 October. LaSalle units now hold first and second places in the world for a continuous run of any LWR (Light Water Reactor).

The world record for any type of reactor is held by a CANDU. This is Pickering unit 7, a 516 MWe (net) unit in Ontario, Canada, with a continuous run of 894 days when it came off-line for maintenance in 1994 October. This unit was put into commercial operation in 1985 January. CANDU is a PHWR (Pressurized Heavy Water Reactor). Rajasthan unit 5, a 202 MWe (net) PHWR in India, put into commercial operation in 2010 February, holds second place to Pickering unit 7 in world ranking after completing a 765 day continuous run and going into its planned biennial maintenance outage in 2014 September. Besides these record breaking runs there have been many runs of over 400 days by the different types of reactors.

These long runs are terminated when it is time for the planned maintenance outage and are not extended until safety targets can no longer be met, which would mean shutting down the unit at an inopportune time. The practical limit of continuous operation of PWRs and BWRs is set by the need to replace about a third of the nuclear fuel and do maintenance after about two years (720 days) or less. In the U.S. most light water reactors units operate on a 18 month fuel cycle and have maintenance outages scheduled for the spring and autumn months when electricity demand is low. Since a pressure tube PHWR like CANDU can refuel on-line at power the length of continuous operation is indeterminate but in practice there is a need to come off-line for certain tests, maintenance and inspections and upgrades that cannot be done at power. For a PHWR the run could be terminated by initiating one of the two reactor safety shutdown systems with the other reactor safety shutdown system being tested during the maintenance outage. The Enhanced CANDU 6 (EC6) is designed to operate for about three years (1080 days) before coming off line for a month for maintenance and inspections. Having some testing and maintenance done on-line reduces the inspection load during unit maintenance outage.

CANDU cousins in India – Performance in 2014

By: Donald Jones, P.Eng., retired nuclear industry engineer

Most of India’s nuclear reactors are of the pressurized heavy water reactor (PHWR) type with horizontal pressure tubes, just like the Canadian designed CANDU. In fact the first PHWR in India was the Rajasthan Atomic Power Project (RAPP) unit and was a CANDU designed by Atomic Energy of Canada Limited (AECL) that used the Douglas Point unit in Ontario as reference design but modified to aid localization. RAPP-1 entered commercial operation 1973 December. While RAPP-1 was being constructed the design of RAPP-2 was started (Author’s note: I know because I was part of design team). However the detonation of a nuclear device by India in 1974 curtailed completion of the design by AECL and India was on its own as far as nuclear technology was concerned. The design was completed by India and RAPP-2 eventually entered commercial operation in 1981 April. Since those early days India has developed its own indigenous designs of PHWRs with net electrical outputs of 202 MW, 490 MW, and 630 MW. They bear little to no relation to Douglas Point. All 18 PHWR units operating in 2014 (including RAPP-1 which has been shutdown since 2004) were 202 MW (220 MW gross) except for two 490 MW (540 MW gross) units. There were four 630 MW (700 MW gross) units under construction with none in operation. All nuclear power units, except for RAPP-1, are designed, owned, and operated by Nuclear Power Corporation of India Ltd. Several of the country’s PHWRs have been refurbished for extended life operation. For more detailed information on the Indian nuclear program see, Nuclear Power in India (reference 1).

CANDU 6 Performance in 2014

By: Donald Jones, P.Eng., retired nuclear industry engineer, 2015 March 24

History

Following on from some early conceptual work by Canadian General Electric (CGE), Atomic Energy of Canada Limited (AECL) based the CANDU 6 design on the four unit Pickering A station (that was brought into service 1971-1973) but as a single unit station with a significant power increase, major equipment simplifications, improvements in shutdown and emergency core cooling systems, extensive use of digital computers for control and safety systems etc. In fact the CANDU 6 is unrecognizable as being based on Pickering except maybe for the fuel channel sizing, even though fewer channels are in CANDU 6, and the two loop primary heat transport system that were retained. Since Ontario Hydro was enamored by multi-unit stations CANDU 6 was intended as a single unit for out of province build including off shore. The two lead CANDU 6 projects were Gentilly 2 in Quebec and Point Lepreau in New Brunswick and these were quickly followed by Embalse in Argentina and Wolsong, now Wolsong 1, in South Korea and all came into service in the early to mid 1980s. These can be regarded as the first tranche of CANDU 6 build.

The second tranche of CANDU 6 units came with Wolsong 2, 3 and 4 in South Korea, Cernavoda 1 and 2 in Romania, and Qinshan 4 and 5 in China (the other units at Qinshan site are not CANDU), all entering service between 1996 to 2007. Each of the second tranche CANDU 6 units incorporate lessons learned from operation of the earlier units with changes to meet latest regulatory codes and standards. All three Wolsong units came in on budget and on schedule and the two Qinshan units came in under budget and ahead of schedule. In fact the total project schedule for the CANDU 6 units at the Qinshan site in China was 81 months from contract effective date to in-service.

Capacity Factor

Unlike the 2013 CANDU 6 performance figures (reference 1) the Capacity Factors are taken from the Power Reactor Information System (PRIS) database of the International Atomic Energy Agency (IAEA). Note that the Load Factor term used in the PRIS database has the same meaning as Capacity Factor. Capacity Factors are based on the (net) Reference Unit Power and on the (net) Electricity Supplied figures, as defined in the PRIS database.

CANDU 6 Units

Performance of Ontario’s CANDU nuclear generating stations in 2014

By: Donald Jones, P.Eng., retired nuclear industry engineer, 2015 March 23

At the end of 2014 Darlington had a four unit average lifetime Capacity Factor (CF) of 84.0 percent and an average annual CF of 91.1 percent. Bruce A had a four unit average lifetime CF of 68.2 percent and an average annual CF of 79.6 percent. Bruce B had a four unit average lifetime CF of 83.5 percent and an average annual CF of 87.3 percent. The six unit Pickering station had a six unit average lifetime CF of 74.5 percent and an average annual CF of 74.4 percent.

Unlike the 2013 performance figures (reference 1) the raw performance data for 2014 are taken from the Power Reactor Information System (PRIS) database of the International Atomic Energy Agency (IAEA). Note that the Load Factor term used in the PRIS database has the same meaning as CF. CFs are based on the (net) Reference Unit Power and on the (net) Electricity Supplied, as defined in the PRIS database.

The performance of some of Ontario’s nuclear generating stations is affected by the surplus of generation in the province. The surplus usually arises because of unreliable intermittent wind generation coming in at times of low demand and wind generation is expected to increase even more over the next several years. Some nuclear units saw electricity output reductions during periods of surplus baseload generation (SBG). This means the CFs are not a true performance indicator for those units (reference 2). A better metric of performance in these cases would be the Unit Capability Factor (UCF – used by Ontario Power Generation and by Bruce Power) or the Energy Availability Factors (EAF) that are shown in the PRIS database. The EAF adjusts the available energy generation for energy losses attributed to plant management and for external energy losses beyond the control of plant management while the UCF only includes energy losses attributed to plant management and excludes the external losses attributed to grid related unavailability and other things. This means that on unreliable grids, for example, UCF will be significantly higher that EAF but for Ontario there will be no significant difference. The UCF and the EAF take into account reductions in plant output due to load cycling and load following. For units that load cycle and/or load follow the CF will be significantly lower than the EAF. For example, Bruce A unit 4 has a 2014 annual CF of 94.3 percent and an EAF of 99.4 percent. The UCF and the EAF are based on reference ambient conditions so, unlike the CF, they cannot exceed 100 percent. The only reason for using the EAF here (see later for Bruce units) instead of the UCF is that EAFs are now available in PRIS and UCFs are not presently available (well, the author could not find them).

South Korea’s first CANDU 6, Wolsong unit 1, is cleared to restart

By: Donald Jones, P.Eng., retired nuclear industry engineer, 2015 February

Wolsong unit 1, a CANDU 6 unit in South Korea operated by the Korea Hydro and Nuclear Power (KHNP) Company, was cleared for continued operation on 2015 February 27 by the Korean nuclear regulator (the Nuclear Safety and Security Commission – NSSC) after being out of service since 2012 November. Wolsong 1 was the first CANDU 6 unit in South Korea (reference 1) and went into commercial operation in 1983 April. The unit was taken out of service for refurbishment in 2009 April. At the end of 2008, the last full year of operation before the shutdown for refurbishment, the annual capacity factor was 93.2 percent and the lifetime capacity factor was 87.0 percent. The capacity factors are taken from the Power Reactor Information System (PRIS) database of the International Atomic Energy Agency (IAEA). Note that the Load Factor term used in the PRIS database has the same meaning as capacity factor. Capacity factors are based on the (net) Reference Unit Power and on the (net) Electricity Supplied, as defined in the PRIS database. For the 5 years prior to the refurbishment outage the average annual capacity factor was 89.4 percent.

CANDU 6 shares the load with light water reactors in China and South Korea

By: Donald Jones, P.Eng., retired nuclear industry engineer – 2014 July

There are two CANDU 6 reactors in China and four in South Korea in a growing sea of pressurized light water reactors. It will be interesting to see how they are performing while operating in the same regulatory environment and safety culture as the other units on the power grid. Performance will be related to the lifetime capacity factor/load factor at the end of year 2013. The International Atomic Energy Agency (IAEA) publishes statistics on all the world’s power reactors in its Power Reactor Information System (PRIS) database.

China

As of the end of 2013 China had 17 operating nuclear power plants, two them of the CANDU 6 type. As of 2014 July China had 20 units in operation with 28 under construction, more about to start construction and more planned. China aims to have 58,000 MW of installed nuclear capacity by 2020, 150,000 MW by 2030 and much more by 2050. For perspective the US has about 100,000 MW installed and Ontario about 13,000 MW. All China’s nuclear units, except for the two French units at Daya Bay and the Chinese designed CNP-300 at Qinshan, entered commercial operation after 2002 and have excellent lifetime capacity factors. The average lifetime capacity factor for the 17 operating units was 88.3 percent. China is now largely self sufficient in power plant design and construction. For anyone interested in China’s massive nuclear plant expansion see, Nuclear Power in China (reference 1).

There are two Russian VVER V-428, each 1060 MWe gross, that went into commercial operation between 2007 May and 2007 August. The average lifetime capacity factor at the end of 2013 was 86.3 percent.

There are four French three loop Framatome (now Areva) units each of around 990 MWe gross that went into commercial operation between 1994 February and 2003 January, the first two units are at Daya Bay with most of their output feeding Hong Kong. The average lifetime capacity factor was 86.4 percent.

China’s four CPR-1000 units each of around 1080 MWe gross are significantly upgraded from Framatome’s Daya Bay and subsequent units and went into service between 2010 September and 2013 June. The average lifetime capacity factor was an excellent 92 percent for these relatively new units. More of these units will be built but not as many as envisaged before the Fukushima event. Instead China is working with Westinghouse in the construction of four, Generation III+, AP1000 units (1,250 MWe gross) with many more to follow and plans to develop its own CAP1000 version and eventually a larger CAP1400.

China’s four CNP-600 units (and one CNP-300) each of around 650 MWe gross are based on early Framatome two loop design work and are all located at the Qinshan site. They went into commercial service between 2002 April and 2011 December. The average lifetime capacity factor was 87.6 percent. China’s first indigenously designed nuclear plant, a two loop CNP-300 unit of 310 MWe gross also at the Qinshan site, had a lifetime capacity factor of 81.1 percent.

The two Canadian CANDU 6 units are each 728 MWe gross at the Qinshan site and went into commercial service between 2002 December and 2003 July. The average lifetime capacity factor was 91.6 percent for these relatively old (for China, anyway) units at the end of 2013. The average capacity factor for the 15 pressurized light water reactors in China was 87.9 percent and if the two CANDU 6 units are included it would be 88.3 percent, an excellent performance by the Chinese nuclear fleet.

South Korea