The primary driver for injecting zinc is the opportunity to obtain dose rate reductions. In other cases, zinc injection is part of a plant’s overall materials reliability programme aimed at mitigating primary water stress corrosion cracking initiation. The Electric Power Research Institute (EPRI) has ongoing research in radiation field management, materials and fuel reliability related to zinc injection. These programmes align with ongoing efforts to increase fuel performance and equipment reliability as well as balancing the radiation field management challenges. This paper is a review of the current industry status related to zinc injection.

The number of pressurized water reactors (PWR) injecting zinc into the primary system has increased from 17 units in 2004 to 73 in 2010, with 10 more planned within the next two years. This represents about 27% of the operating PWRs in the world and 56% of the PWRs in the U.S. By the end of 2011 the percentage of PWRs injecting zinc is expected to increase to approximately 31% worldwide. Plants inject zinc for two primary reasons: dose rate reduction and primary water stress cracking corrosion (PWSCC) initiation mitigation. Additional goals for zinc injection programmes are to mitigate corrosion product generation and crud deposition on fuel surfaces. Approximately 85% of the plants injecting zinc report dose rate reduction as the primary goal, with the remainder identifying PWSCC mitigation and crud mitigation as the primary drivers. Typical zinc injection strategies employ reactor coolant system (RCS) zinc concentrations of 5-20 ppb. Additional plant experience with zinc concentrations as high as 35-40 ppb has also been obtained.

Zinc injection achieves the noted benefits via mechanisms at the molecular level. As zinc is incorporated into the oxide films of wetted surfaces in an operating PWR, it changes the morphology and composition of oxide films, thereby changing their corrosion characteristics. In addition, it is believed that zinc displaces nickel and cobalt from the crystalline lattice sites in the inner layer of system surfaces. With time, this process makes the oxide layers thinner, more stable and more protective.

Zinc impact on PWR primary chemistry

Farley Unit 2 was the first plant to inject zinc starting in 1994. Farley used naturally-occurring zinc acetate and targeted an RCS zinc concentration of 35-40 ppb. Naturally-occurring zinc consists of 48.6% Zn-64, 27.9% Zn-66, 4.1% Zn-67, 18.8% Zn-68, and 0.6% Zn-70 [3]. In the presence of a neutron flux, Zn-64 can absorb a neutron to become Zn-65, which is radioactive (1.1 MeV gamma) with a 243.8 day half-life. Therefore, plants adding natural zinc experience a smaller radiation dose benefit because of the production of Zn-65. Similarly high target zinc concentration programmes were implemented in the U.S. at Farley 1, Diablo Canyon 1 and 2 and Beaver Valley 1. In the mid-1990s, the large difference in cost between natural and depleted zinc favoured injection programmes using natural zinc.

Subsequent zinc injection projects using depleted zinc demonstrated that additional dose rate reductions could be achieved. Depleted zinc does not introduce Zn-64 into the coolant, thereby avoiding the creation of Zn-65 and enabling greater dose rate reductions. To date all of the U.S. plants that once employed natural zinc have transitioned to depleted zinc, a shift further supported by the significant cost reduction in depleted zinc. Additionally, all of these plants noted a reduction, or levelling, in system zinc demand and an increase in the contribution of 65Zn to the dose-impacting radioisotopes while still using natural zinc.

A significant amount of RCS chemistry data has been collected as part of the utility-specific zinc programmes over the past several years. The EPRI PWR Zinc Users Group sponsored research to assess the impact of RCS zinc addition on nickel and radiocobalt concentrations during both operating and shutdown periods to determine if plant responses to initial zinc injection could be predicted.

Nickel

A principal concern regarding the plant response to initial injection is that dissolved zinc will interact with ex-core oxide films in a manner that releases nickel into the primary coolant system. Nickel released by this mechanism could deposit in the core and challenge fuel performance. Primary system chemistry data (principally nickel concentrations and radiocobalt activities) were evaluated for the cycles in which zinc was first injected. Assessments included comparisons of concentrations and activities before and after zinc injection as well as comparison of these periods to similar times in previous cycles. The mass of nickel released during shutdown was also assessed [4].

  • The assessments did not discern a statistically significant increase in coolant nickel concentration upon initial zinc injection. However, the statistical evaluations of data from several units evaluated were inconclusive in this regard; that is, there could have been a nickel response at some low level but it could not be validated with the statistical tests used.
  • There was no statistically significant increase in the mass of nickel released during shutdown chemistry manoeuvres following the initial zinc injection (see Table 1).

Radiocobalts

The impact of the first injection of zinc on primary coolant radiocobalt activities was assessed using raw data from several units and using published data from several others. Table 2 tallies the results. As these tables indicate, radiocobalt responses have been observed at some units, but not others. The analysis could not identify a factor that could be correlated to whether or not a unit showed a radiocobalt response [4].

Figures 2 through 4 illustrate the variety of plant responses observed in Co-58 associated with RCS zinc addition [5]. Figure 2 depicts one pre-zinc cycle (Cycle 11) and all subsequent data for Co-58 values for Byron 2. Figure 3 depicts one pre-zinc cycle (Cycle 11) and all subsequent data for Co-58 values for Braidwood 2. Byron 2 and Braidwood 2 are essentially sister plants because they have comparable core design and core management strategies and have closely mirrored each other with respect to primary chemistry programmes. As such, it would be expected that both plants would have relatively similar responses to zinc addition. Figure 2 illustrates essentially no impact on Co-58 values for the initial cycle of zinc addition at Byron 2 while Figure 3 illustrates a strong response (increase) in Co-58 levels for the initial cycle of zinc addition at Braidwood 2. Figure 4 depicts the Three Mile Island 1 Co-58 values for one pre-zinc cycle (Cycle 15) and all subsequent data for Co-58 values. Although the Co-58 values were already trending upward prior to initiating zinc injection, the nearly 1000-fold increase was statistically significant compared to Cycle 15.

Regular updates to these evaluations to assess the longer-term impacts of zinc injection on RCS nickel and radiocobalt concentrations are planned. These results, while limited to the impacts identified during the initial zinc injection cycle, indicate that further assessment is warranted of the mechanism by which zinc is incorporated into the RCS oxide layers. Additionally, this data suggests that while an increase in radiocobalt levels may not occur in conjunction with zinc injection, one should be planned for.

Zinc impact on fuel

EPRI has an extensive programme in place to ensure fuel integrity and performance are not challenged by zinc injection. Based on fuel surveillance programmes at six plants with increasing fuel duty, EPRI has made several observations to date:

1. Zinc has not caused an increase in fuel cladding corrosion at any of the EPRI- sponsored campaigns or other reported campaigns

2. No abnormal buildup of crud has been observed

3. No fuel performance issues (such as axial offset anomaly) have been reported that were directly related to zinc [6].

The experience base for high-duty plants injecting zinc continues to grow. There are currently eight high-duty plants injecting zinc that have done so successfully, obtaining dose rate reduction and materials benefits, with no reported fuel concerns. For high-duty plants, a risk assessment specific to the cycle and zinc injection strategy is an important part of the overall zinc injection program. Byron 1 and Braidwood 1, the two highest-duty four-loop PWRs in the U.S., successfully implemented zinc addition in 2010. The first cycle of injection at Braidwood 1 concluded in the autumn of 2010 with no reported negative fuel impacts [7]. The first cycle of injection at Byron 1 will conclude in the spring of 2011.

Plants injecting zinc have used a combination of fuel cladding, including zircaloy-4, ZIRLO, low-tin zircaloy, M5, OPTIN and various combinations of these claddings. Industry exposure for the three highest exposure claddings are shown in Table 3, with ZIRLO cladding having the highest exposure [1,8].

Zinc impact on PWR dose rates

As part of the EPRI standard radiation monitoring and chemistry monitoring and assessment programmes, refuelling outage dose rate data, and startup, shutdown and operating plant chemistry data is collected for plants worldwide. Trending of refuelling outage dose rate data and shutdown chemistry releases continues to show trends of reduced dose rates and releases for plants injecting zinc.

The 2003 EPRI report, PWR Operating Experience with Zinc Addition and the Impact on Plant Radiation Fields (TR-1003389), initially developed correlations between zinc exposure and dose rate reductions as measured at standardized survey points for plants with Alloy 600, Alloy 690 and Alloy 800 steam generator tubing material. These correlations were updated for the 2006 EPRI Pressurized Water Reactor Zinc Application Guidelines (TR-1013420) and again for revision 6 of the EPRI Pressurized Water Reactor Primary Water Chemistry Guidelines (TR-1021133), published in late 2010.

Figure 6 plots the dataset of cumulative dose reductions to zinc exposure for the different steam generator (SG) tubing materials. It is logarithmically fitted [2]. These correlations have been used to predict potential dose rate reductions resulting from specific zinc programmes and to assess the efficacy of a completed zinc injection cycle.

It should be noted that initially Alloy 600 and 690 behave slightly differently and these curves do not take into account SG tubing manufacturing processes or release rates observed in these different materials. These dose rate reduction correlation curves are in the process of being revised to support the in-process revision of the PWR zinc application guidelines.

Figure 7 shows a compilation of channel head dose rates for plants injecting zinc [9]. The reduction in both hot leg and cold leg dose rates were achieved as a result of zinc injection at plants with Alloy 690TT, and Alloy 600MA and 600TT steam generator tubing.

Figure 8 shows dose trends during refuelling outages at the Callaway plant, which initiated zinc addition in Cycle 13 and continues to inject zinc presently.

Byron Unit 2, a high-duty four-loop PWR, implemented zinc addition during the last six months of cycle 12 and has seen continued dose rate reductions. Table 4 shows dose rate trends for three cycles prior to zinc implementation through cycle 15 (the most recently reported cycle) [5]. Byron 2 has experienced a 67% dose rate reduction between refuelling outage 11 and refuelling outage 15. Exelon attributes approximately 42% of this reduction to have occurred as a result of the zinc addition, with the balance attributed to end-of-cycle boron levels [11]. In summary, reductions in shutdown dose rates have been routinely observed at PWRs following zinc addition.

Conclusions

Zinc continues to be the most consistent option for dose rate reduction. A complete understanding of plant materials, core design and chemistry issues is essential prior to commencing a zinc injection programme. Consistent with EPRI Pressurized Water Reactor Primary Water Chemistry Guidelines, a nuclear utility’s primary chemistry strategic plan should, as a minimum, review and document the bases for injecting zinc or not.