ML13098A431
| ML13098A431 | |
| Person / Time | |
|---|---|
| Site: | South Texas  |
| Issue date: | 04/08/2013 |
| From: | Sande T Enercon Services |
| To: | Plant Licensing Branch IV |
| Singal B | |
| Shared Package | |
| ML13098A368 | List: |
| References | |
| TAC MF0613, TAC MF0614 STP‐RIGSI191‐V03.6.1, Rev 0 | |
| Download: ML13098A431 (8) | |
Text
South Texas Project RiskInformed GSI191 Evaluation Development of Chemical Effects Modules for Risk Informed GSI191 Resolution Document: STPRIGSI191V03.6.1 Revision: 0 Date: March 28, 2013 Prepared by:
Timothy D. Sande, Enercon Services, Inc.
Reviewed by:
Blake Stair, Enercon Services, Inc.
Janet Leavitt, University of New Mexico Kerry Howe, University of New Mexico Zahra Mohaghegh, University of Illinois Seyed Reihani, University of Illinois Approved by:
Ernie Kee, South Texas Project
South Texas Project RiskInformed GSI191 Evaluation Development of Chemical Effects Modules STPRIGSI191V03.6.1 Revision 0 Purpose The purpose for this document is to provide a description of the generic methodology for evaluating the impact of chemical effects within a riskinformed framework. Although the approach is generic, all four of the chemical effects modules that are described must be developed based on plantspecific parameters. The description of the modules includes high level discussion of the types of testing and analysis required for each module. However, the numerous details required for fully developing and implementing each module will be described in other documents.
Introduction To evaluate the possible outcomes should a loss of coolant accident (LOCA) occur, the riskinformed approach for resolving GSI191 relies on an evaluation of thousands of scenarios (different break sizes and locations, variations in debris characteristics, ranges of possible water levels and flow rates, etc.).
Each input has a range of possible values dependent on both random and systematic variations. The input variable probability distributions are thoroughly evaluated using statistical sampling methods.
However, physical models or test data are necessary to understand the outcome of a given set of conditions.
For the chemical effects portion of the GSI191 evaluation, the input variables include timedependent pool temperature and pH, insulation and coatings debris quantities, exposed surface areas for concrete and reactive metals in containment, initial chemistry of the RCS and RWST, buffer type and quantity, and other factors. The goal of the chemical effects evaluation is to quantify the effects of insoluble chemical products1 on debris bed head loss. This can be difficult because chemical products won't necessarily form in all scenarios. Also, many of the input parameters have competing effects that tend to offset each other. For example, maximizing the temperature profile will tend to maximize the quantity of aluminum released into solution, but will also raise the solubility limit for aluminum precipitates to form.
For deterministic evaluations, the approach that has typically been used by the industry (as described in WCAP16530NP) is to calculate the quantity of materials released into solution based on bounding input conditions and to assume full precipitation of aluminum (and in some cases, calcium and silicon) to determine the total quantity of precipitates for a given scenario. The head loss induced by these precipitates is determined by performing head loss tests with surrogate mixtures that are recognized as a very conservative representation of the actual precipitates that may form.
1 The term chemical products is used in this document to refer to insoluble products that may result from precipitation in the bulk solution, formation of scale on metal surfaces, nucleation and growth of crystals on fiber surfaces, or any other formation mechanism.
Page 2 of 8
South Texas Project RiskInformed GSI191 Evaluation Development of Chemical Effects Modules STPRIGSI191V03.6.1 Revision 0 Calculating the impact of chemical effects over the wide range of possible postLOCA conditions requires the development of four chemical effects modules: (1) a solubility limit calculator to determine the concentration where a chemical product could form as a function of temperature and pH; (2) a chemical release module to determine the timedependent concentration of important chemical constituents based on corrosion or dissolution release rates, material quantities, and temperature and pH profiles; (3) a product formation module to determine the type and quantity of chemical products, the location where the products would form, and the characteristic morphology and size of the products; and (4) a chemical head loss module to determine the effect of the products on debris bed head loss. These modules are described in more detail below.
Module 1: Solubility Limits The solubility limit is the aqueous concentration required for a given product to form. This limit varies depending on the type and morphology of the product. In general, the solubility limit for a given chemical product decreases with decreasing temperature. However, some chemical products exhibit retrograde solubility, which means that the solubility limit increases with decreasing temperature. The solubility limit is also dependent on the pH, and there is generally a unique pH value that minimizes a chemical products solubility limit for any given temperature (i.e., the solubility limit increases as the solution becomes more acidic or basic).
Based on extensive testing within the chemical industry, equilibrium conditions have been established for a wide range of chemical products. Using the existing databases and current literature, the solubility limits (based on equilibrium conditions) for relevant plantspecific products can be estimated with thermodynamic modeling. As with other aspects of the riskinformed evaluation, the thermodynamic model predictions include a level of uncertainty (due to uncertainties in the input variables) that must be considered in the evaluation. Figure 1 illustrates the solubility limit (with uncertainty bands) for a chemical product as a function of temperature.
Page 3 of 8
South Texas Project RiskInformed GSI191 Evaluation Development of Chemical Effects Modules STPRIGSI191V03.6.1 Revision 0 Figure 1: Illustration of aluminum solubility limit versus temperature at a constant pH The key for developing this module is to identify the chemical products that are most likely to form under plantspecific conditions. For South Texas Project (STP), which uses trisodium phosphate (TSP) as a buffering agent, the most likely products are considered to be aluminum hydroxide (Al(OH)3) or other aluminum products, calcium phosphate (Ca3(PO4)2), and zinc phosphate (Zn3(PO4)24H2O). Although the thermodynamic models are based on extensive testing, additional benchtop testing may be required to confirm the potential chemical products that will be evaluated for Module 1.
Module 2: Corrosion and Dissolution Release The second module provides a prediction of the timedependent release of various materials (Al, Ca, Si, Zn, etc.) into solution based on the corrosion/dissolution release rates, exposed quantity or surface area of contributing materials (e.g., insulation and coatings debris and metals in containment), and containment conditions. This module is essentially equivalent to the WCAP16530NP calculator.
However, some modifications may be necessary to more accurately calculate timedependent material release for plantspecific conditions. The potential modifications include:
An adjustment to the aluminum release rate to address NRC concerns that the WCAP calculator underpredicts the rate (although not the total quantity released over 30 days) by approximately a factor of 2.
Incorporation of a zinc corrosion rate (i.e., for galvanized steel and inorganic zinc coatings).
Page 4 of 8
South Texas Project RiskInformed GSI191 Evaluation Development of Chemical Effects Modules STPRIGSI191V03.6.1 Revision 0 Incorporation of the inhibition of aluminum corrosion by silicon and phosphate based on WCAP 16785NP.
Incorporation of the inhibition of aluminum corrosion by zinc.
As with the solubility module, uncertainty in the corrosion and dissolution rates must be considered in the evaluation. Figure 2 illustrates the time and temperaturedependent concentration that would be determined from Module 2 and the temperaturedependent solubility limit that would be determined from Module 1 (including uncertainties). The point where the two lines cross (i.e., where the concentration exceeds the solubility limit) is representative of the time when a chemical product could first begin to form. Note that since containment pool temperature generally decreases over time, the solubility curve shown in Figure 1 is inverted in Figure 2.
Figure 2: Comparison of timedependent solubility limit and aqueous concentration The majority of the release module is based on existing test data documented in WCAP16530NP and WCAP16785NP. However, additional benchtop corrosion tests may be required to address the corrosion of galvanized steel and IOZ coatings as well as the inhibitory effects of zinc on aluminum corrosion.
Page 5 of 8
South Texas Project RiskInformed GSI191 Evaluation Development of Chemical Effects Modules STPRIGSI191V03.6.1 Revision 0 Module 3: Product Formation The product formation module is intended to predict the type, quantity, location, morphology, size, and transport of the chemical products. The product types (e.g., aluminum products or zinc phosphate) can be determined based on whether the concentration exceeds the solubility limit. Similarly, the quantity (including uncertainty) can be calculated based on a chemical mass balance from the results of the first two modules.
Chemical products could form at the following locations: (1) precipitation within the bulk solution, (2) the buildup of scale on material surfaces, or (3) growth within the debris bed itself. As long as a significant quantity of chemical product forms, precipitation within the bulk solution can be readily identified by taking water samples during a test and measuring turbidity or comparing the chemical concentrations of filtered and unfiltered samples. For example, if the aluminum concentration in an unfiltered sample is significantly higher than in a filtered sample, it is a good indication that an aluminum precipitate has formed in the bulk solution. If a significant quantity of chemical product forms as a scale on metal surfaces within a test, this can be observed by comparing the surfaces before and after the test. For example, during the 2012 large break CHLE test, a significant quantity of zinc phosphate formed on the galvanized steel coupons. If the chemical products form within the debris bed itself, it may be difficult to directly observe the products at the end of the test, and the filtered and unfiltered water samples would be approximately the same. However, significant product growth within the debris bed can be indirectly observed by otherwise unexplained increases in the debris bed head loss.
The morphology and size can be determined if the quantity of the product formed is sufficient to permit its collection and characterization. The morphology is important because an amorphous precipitate may have significantly different head loss characteristics (and would generally be more detrimental) than a crystalline product. The size is also important because precipitates that form in the bulk solution could pass through a debris bed if they are substantially smaller than the void space dimensions of the bed, whereas precipitates that are much larger could be readily captured by a debris bed.
The transport for chemical products depends on the location where the products are formed as well as the size and morphology. If the products form within the debris bed itself, the transport is irrelevant.
However, if the products form on metal surfaces, they may or may not detach from the surfaces and transport to the debris bed.
The key to developing Module 3 is to conduct tests in which the quantity of product formation is large enough to observe the location where it occurs and to collect and characterize the product. As discussed above, three types of chemical products (aluminum, calcium, or zinc products) could potentially form for the full range of STP conditions. A significant quantity of zinc phosphate was observed to form as a crystalline product on the submerged galvanized steel coupons and bags of zinc granules in the STP Page 6 of 8
South Texas Project RiskInformed GSI191 Evaluation Development of Chemical Effects Modules STPRIGSI191V03.6.1 Revision 0 largebreak CHLE test. A small portion of the zinc product detached from the surfaces where it formed and settled on the floor of the CHLE tank or transported to the debris beds. No significant quantities of aluminum or calcium products were observed in this test, however. Therefore, additional testing is required to identify the formation locations and characteristics of aluminum products and calcium phosphate products unless testing and Modules 1 and 2 demonstrate that a product will never form under the range of plantspecific conditions. To accomplish this, the tests should be designed to encourage the potential formation of each of these products. This is one purpose for the two 10day tests described in the current CHLE test plan.
Module 4: Chemical Head Loss The purpose of the chemical head loss module is to quantify the overall effects of chemical product formation. In some cases, the first three modules may predict that chemical products would not form or that the chemical products that form would not be transported to the debris bed. For these cases, there would be no chemical head loss. However, in cases where a significant quantity of chemical products form and transport to the debris bed (or grow within the debris bed), the increase in debris bed head loss must be determined. If the morphology of the chemical products is crystalline, it may be possible to treat it similar to particulate debris within the conventional debris head loss correlation. However, amorphous precipitates do not behave like conventional particulate debris and generally have a much greater impact on head loss. The change in head loss due to chemical products is also dependent on the existing conventional debris bed. A thin bed of fiber debris with negligible quantities of particulate would likely have a significantly different response to a given quantity of chemical products than would a thick fiber bed with large quantities of particulate.
The approach that is proposed for quantifying the effects of chemical products on head loss is to develop a bumpup factor correlation for the range of plantspecific conditions. This differs from previous attempts to develop a universal bumpup factor. Instead, a range of bumpup factors will be determined that correspond to a variety of chemical conditions and conventional fiberglass debris loads.
The testing required to develop Module 4 includes a series of vertical loop tests with chemical conditions focused on the more problematic scenarios (i.e., cases where maximum quantities of chemical products are predicted based on Modules 13) and that consider variations in debris beds (i.e.,
combinations of thin and thick fiber beds with low and high particulate quantities).
Conclusions To implement a riskinformed GSI191 resolution approach, four chemical effects modules are required.
The first module will be used to determine the solubility limit as a function of temperature and pH for potential chemical products that may form. This module will be based on thermodynamic modeling, although testing may be required to verify the module. The second module will be used to predict the timedependent concentration of various chemicals as a function of the quantity and surface area of Page 7 of 8
South Texas Project RiskInformed GSI191 Evaluation Development of Chemical Effects Modules STPRIGSI191V03.6.1 Revision 0 various materials and of the temperature and pH. This module will be based primarily on the corrosion/dissolution equations in WCAP16530NP, although some additional benchtop or integrated tank tests may be necessary.
The first two modules feed into the third module, which will be used to determine the type, quantity, location, and characteristics of the chemical products. The type and quantity of product will be determined by comparing the results of Modules 1 and 2, and the location and characteristics of the products will be determined based on integrated tank tests where the products are observed to form.
The fourth module will quantify the head loss due to the chemical products predicted to form by the third module. The chemical head loss module will be based on integrated vertical loop testing with prototypical debris beds where the increase in head loss can be measured as a function of the chemical products formed and the debris bed where the products are collected.
Once all of these modules are fully developed, it will be possible to fully evaluate chemical effects for the range of plantspecific conditions. At this point, it may be beneficial to conduct a 30day integrated test for the limiting scenario (as predicted by the four modules) to confirm that the modules provide a reasonable prediction.
Page 8 of 8