Search NASA⌕ Search

Engineering topics

Richard Dimenna, R

Publications and source records attributed to Richard Dimenna, R.

TANK 32 EVAPORATOR FEED PUMP TRANSFER ANALYSIS

The transfer of liquid salt solution from Tank 32 to an evaporator is to be accomplished by activating the evaporator feed pump, with the supernate surface at a minimum height of approximately 74.4 inches above the sludge layer, while simultaneously turning on the downcomer with a flow rate of 110 gpm. Previously, activation of the evaporator feed pump was an isolated event without any other components running at the same time. An analysis of the dissolved solution transfer has been performed using computational fluid dynamics (CFD) methods to determine the amount of entrained sludge solids pumped out of the tank toward the evaporator with the downcomer turned on. The analysis results shows that, for the minimum tank liquid level of 105 inches above the tank bottom (which corresponds to a liquid depth of 74.4 inches above the sludge layer), the evaporator feed pump will contain less than 0.1 wt% sludge solids in the discharge stream, which is an order of magnitude less than the 1.0 wt% undissolved solids (UDS) loading criteria to feed the evaporator. Lower liquid levels with respect to the sludge layer will result in higher amounts of sludge entrainment due to the increased plunging jet velocity from the downcomer disturbing the sludge layer.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

ADVANCED MIXING MODELS

The process of recovering the waste in storage tanks at the Savannah River Site (SRS) typically requires mixing the contents of the tank with one to four dual-nozzle jet mixers located within the tank. The typical criteria to establish a mixed condition in a tank are based on the number of pumps in operation and the time duration of operation. To ensure that a mixed condition is achieved, operating times are set conservatively long. This approach results in high operational costs because of the long mixing times and high maintenance and repair costs for the same reason. A significant reduction in both of these costs might be realized by reducing the required mixing time based on calculating a reliable indicator of mixing with a suitably validated computer code. The work described in this report establishes the basis for further development of the theory leading to the identified mixing indicators, the benchmark analyses demonstrating their consistency with widely accepted correlations, and the application of those indicators to SRS waste tanks to provide a better, physically based estimate of the required mixing time. Waste storage tanks at SRS contain settled sludge which varies in height from zero to 10 ft. The sludge has been characterized and modeled as micron-sized solids, typically 1 to 5 microns, at weight fractions as high as 20 to 30 wt%, specific gravities to 1.4, and viscosities up to 64 cp during motion. The sludge is suspended and mixed through the use of submersible slurry jet pumps. To suspend settled sludge, water is added to the tank as a slurry medium and stirred with the jet pump. Although there is considerable technical literature on mixing and solid suspension in agitated tanks, very little literature has been published on jet mixing in a large-scale tank. If shorter mixing times can be shown to support Defense Waste Processing Facility (DWPF) or other feed requirements, longer pump lifetimes can be achieved with associated operational cost and schedule savings. The focus of the present work is to establish mixing criteria associated with the waste processing at SRS and to quantify the mixing time required to suspend sludge particles with the submersible jet pump. Literature results for a turbulent jet flow are reviewed briefly, since the decay of the axial jet velocity and the evolution of the jet flow patterns are important phenomena affecting sludge suspension and mixing operations. One of the main objectives in the waste processing is to provide the DWPF a uniform slurry composition at a certain weight percentage (typically {approx}13 wt%) over an extended period of time. In preparation of the sludge for slurrying to DWPF, several important questions have been raised with regard to sludge suspension and mixing of the solid suspension in the bulk of the tank: (1) How much time is required to prepare a slurry with a uniform solid composition for DWPF? (2) How long will it take to suspend and mix the sludge for uniform composition in any particular waste tank? (3) What are good mixing indicators to answer the questions concerning sludge mixing stated above in a general fashion applicable to any waste tank/slurry pump geometry and fluid/sludge combination? Grenville and Tilton (1996) investigated the mixing process by giving a pulse of tracer (electrolyte) through the submersible jet nozzle and by monitoring the conductivity at three locations within the cylindrical tank. They proposed that the mixing process was controlled by the turbulent kinetic energy dissipation rate in the region far away from the jet entrance. They took the energy dissipation rates in the regions remote from the nozzle to be proportional to jet velocity and jet diameter at that location. The reduction in the jet velocity was taken to be proportional to the nozzle velocity and distance from the nozzle. Based on their analysis, a correlation was proposed. The proposed correlation was shown to be valid over a wide range of Reynolds numbers (50,000 to 300,000) with a relative standard deviation of {+-} 11.83%. An improved correlation including the effect of circulation time was proposed by Grenville and Tilton [11] via a better fit of mixing time data for turbulent jet mixing under a wider range of jet Reynolds numbers (50,000 to 300,000). The circulation time was defined as the liquid volume divided by the entrained flow rate. They assumed that the mixing rate at the end of the jet length controls the mixing time for the entire tank by estimating the kinetic energy dissipation rate as discussed earlier. They predicted that for a given volume, an optimum geometry exists for a mixing vessel, allowing a desired mixing time to be achieved for a minimum power input. The current work will compare their correlation of the jet mixing time with CFD modeling results for their experimental tanks in an attempt to achieve a fundamental understanding of the turbulent jet mixing and to establish mixing indicators.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

F-AREA PUMP TANK 1 MIXING ANALYSIS

The F-area pump tanks are used to transfer supernate, sludge, and other materials. In any transfer, the solution must stay well mixed without allowing particulate matter to settle out of the liquid and, thus, accumulate in the bottom of the pump tank. Recently, the pulse jet mixing in F-area Pump Tank 1 (FPT1) has been decommissioned. An analysis of the liquid transfer through FPT1 has been performed using computational fluid dynamics (CFD) methods to assess whether or not the velocities throughout the tank will remain high enough to keep all particulate suspended using only transfer and recirculation pumps. The following paragraph is an abbreviated synopsis of the transfer procedure for FPT1 [1, 2]. Prior to a transfer, FPT1 begins to be filled with inhibited water through the inlet transfer line (TI). When the tank liquid level reaches 52.5 inches above the absolute tank bottom, the recirculation pump (RI and RO) is activated. At a tank liquid level of 72.5 inches above the absolute tank bottom, the outlet transfer line (TO) is activated to reduce the liquid level in FPT1 and transfer inhibited water to H-area Pump Tank 7 (HPT7). The liquid level is reduced down to 39.5 inches, with an allowable range from 37.5 to 41.5 inches above the absolute tank bottom. HPT7 goes through a similar procedure as FPT1 until both have tank liquid levels of approximately 39.5 inches above the absolute tank bottom. The transfer of inhibited water continues until a steady-state has been reached in both pump tanks. At this point, the supernate/sludge transfer begins with a minimum flow rate of 70 gpm and an average flow rate of 150 gpm. After the transfer is complete, the pump tanks (both FPT1 and HPT7) are pumped down to between 20.5 and 22.5 inches (above absolute bottom) and then flushed with 25,000 gallons of inhibited water to remove any possible sludge heal. After the flushing, the pump tanks are emptied. Note that the tank liquid level is measured using diptubes. Figure 2.1 provides a simplified sketch (not to scale) of FPT1 during the steady-state transfer condition, which consists of two inlet flows that impact the liquid surface as plunging jets and two outlet flows drawn from near the bottom of the tank. During the transfer, the supernate level is held at 39.5 inches above the absolute bottom of the tank [1, 2]. In addition, the FPT1 can contain up to 16.7 wt.% sludge particles within the supernate for a given transfer [2]. Test results from Tank 40 sludge Batch 3 [3] provide a typical range of particulate diameters between 0.1 and 25 {micro}m, with approximately 20 vol.% of the sludge distribution consisting of particles less than 1 {micro}m in diameter. The purpose of this analysis is to estimate FPT1 flow field during the steady-state transfer conditions to ensure that the tank remains mixed and that the velocities throughout the tank are sufficient to keep all sludge particulate suspended.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗