Concentration, Storage and Transfer Engineering (CSTE) requested that the Savannah River Technology Center (SRTC) investigate the behavior of tri-n-butyl phosphate (TPB) and its degradation products (dibutyl phosphate (DBP), monobutyl phosphate (MBP), and butanol) under simulated waste conditions. SRTC also reviewed the impact of other organic materials that are present in small quantities in the high level waste system.

The SRTC evaluation included the following areas:

• TBP hydrolysis. The purpose of this work was to determine the TBP decomposition rate (n-butanol production rate) as a function of temperature, salt solution type, and catalyst.
• TBP, DBP, and n-butanol solubility. The purpose of this work was to help determine the TBP hydrolysis rate and to determine whether DBP and n-butanol would form a floating organic phase in tanks.
• TBP mass transfer rate. Determine the TBP mass transfer rate from a floating organic layer to the aqueous phase. The goal was to show that the TBP hydrolysis process is mass transfer limited and that the butanol production rate is lower than predicted by the hydrolysis data.
• Irradiation of TBP. Irradiate samples of TBP in salt solution and analyze for volatile and semivolatile organic products.
• Evaporator vapor composition. Determine vapor composition during evaporator startup an0...d shutdown in order to determine whether the evaporator passes through a window in which it could exceed the butanol LFL.
• Butanol vapor-liquid equilibrium. Perform n-butanol vapor-liquid equilibrium tests to determine whether the vapor in equilibrium with n-butanol at its solubility exceeds the LFL and to determine whether a floating organic layer reduces the n-butanol vapor pressure.
• Collecting and analyzing Tank Farm samples.
• Determining waste tank and pump tank organic constituent auto-ignition temperatures.
• Evaluating pump tank agitation methods.

The conclusions from this work are the following:

• The TBP measured hydrolysis rates at 40°C, 70°C, and 110°C are 0.023 – 0.030 mg/L hr, 0.26 – 0.38 mg/L hr, and 3.5 – 9.7 mg/L hr, respectively. The measured TBP hydrolysis rates at 30°C and 150°C are 0.025 mg/L hr and 51.6 mg/L hr, respectively. The measured TBP hydrolysis rate at 70°C in dilute salt solution (0.48 M Na) is 1.05 mg/L hr.
• The rate of DBP hydrolysis in alkaline solution is very slow or nonexistent.
• DBP is very soluble in high hydroxide and low hydroxide simulants. It should be transported with the salt in the High Level Waste System and not accumulate as a floating organic phase.
• N-butanol is very soluble in high and low hydroxide simulants. It should not accumulate as a floating layer in waste tanks and pump tanks.
• The TBP hydrolysis process can be mass transfer limited in unagitated waste tanks and pump tanks. If tank conditions cause the TBP decomposition process to be mass transfer-limited, the butanol generation rate will be less than predicted by the hydrolysis experiments.
• The n-butanol vapor-liquid equilibrium test data suggest that a floating organic layer will reduce the n-butanol vapor pressure.
• Analysis of irradiated TBP samples indicated no n-butanol or DBP was formed.

• Analysis of samples from a number of waste and pump tanks indicated the concentration of organic species to be extremely low in all samples. The worst case was Pump Tank 3F where 230 mg/L TBP and 77 mg/L n-paraffin were found; the concentrations of the remaining organics in the tanks were in the low mg/L range. No organic species were found in most vapor samples, and only nanogram/L quantities of organics were found in a few of the vapor samples.

• The TBP hydrolysis model given in this report may be used to calculate TBP hydrolysis rates (and also butanol generation rates) in High Level Waste Tanks as a function of temperature.

• The mass transfer data may be used along with the hydrolysis data to calculate butanol generation rates at temperatures and tank conditions under which the process is mass transfer limited (i.e., high temperatures).

• The hydrolysis data indicate that the rate of DBP hydrolysis in alkaline solutions is very slow or non-existent in the time period of the tests. This supports the assumption that only one butanol molecule is produced by the decomposition of TBP in this time frame. The actual yield of DBP and butanol is somewhat different than predicted by theory, but the concentrations do not change significantly during the latter stages of the tests. The variability of the yields is attributed to analytical uncertainty.

• The solubility tests show that DBP is very soluble in high and low hydroxide salt solutions and should not form a floating organic phase in waste tanks and pump tanks.

• Because of its high solubility and volatility, butanol should not accumulate as a separate floating layer in High Level Waste Tanks.

• TBP has a low solubility and if present could accumulate as separate floating layer. Samples collected to date do not indicate a particular concern with the accumulation of TBP.

• The equilibrium tests indicate that a floating organic layer could reduce the partial pressure of butanol. However, the test method was inconclusive on the magnitude of this effect, and therefore the data generated should not be used quantitatively.

• The results of the evaporator tests are inconclusive and do not provide sufficient evidence to draw any conclusions as to flammability concerns in the evaporator during heat-up and cool-down.

The High Level Waste Tank Farms store and process high-level liquid wastes from a number of sources including F- and H-Canyon, Receiving Basin for Offsite Fuels (RBOF) and the Defense Waste Processing Facility (DWPF). These wastes are made alkaline prior to transfer to the Tank Farm and are subject to acceptance based on their composition. These wastes may contain minor concentrations of organic compounds. The Authorization Basis for the Tank Farms identifies several controls to prevent unwarranted, adverse chemical reactions.¹ However, current analysis of the accident scenarios does not evaluate the impact of the presence of organic compounds. A Potential Inadequacy in the Safety Analysis (PISA) was declared regarding the issue of organic compounds in the waste.² Processing vessels of concern include the pump tanks (PT), waste tanks, (WT), and evaporators.

In particular, the F Canyon personnel routinely flush their evaporators to eliminate safety concerns surrounding hold-up of tributyl phosphate. Canyon personnel ensure the solutions contain less than 0.5 volume % (on average) organic components and discharge the solutions to the Tank Farms. The fate of the tri-n-butyl phosphate (TPB) and the degradation products (dibutyl phosphate (DBP), monobutyl phosphate (MBP), and butanol) are not completely understood. Thus, the concern over flammability issues has remained in the Tank Farm. Concentration, Storage and Transfer Engineering (CSTE) requested that the Savannah River Technology Center (SRTC) investigate the behavior of these organic phosphates under simulated waste conditions.³ Task Technical and Quality Assurance plans approved by CSTE were written to cover work under this program.^4,5,6

In addition to TBP and its decomposition products, there are a number of miscellaneous organic materials added in small quantities through various Tank Farm input streams. These could also impact flammability in the Tank Farm, and were reviewed.

TBP is both difficult to extract from very high salt/high caustic aqueous solutions and for the purposes of this experimental program is present in low concentrations. The analytical requirements of this study necessitated the employment of new strategies for TBP extraction. Two methods were proposed and investigated for this study: a dilution method and the method of standard additions. A dilution study was performed to examine the extraction efficiency of TBP in varying salt solutions. The method of standard additions was employed to more accurately evaluate low level TBP concentrations.

The method of standard additions was employed for all samples in this study. In this technique, the sample is analyzed twice. First the sample is analyzed as is; then a known amount of the analyte (TBP) is added and the sample is re-analyzed. The difference in the analyses reflects the amount of analyte known to have been added and allows more accurate assessment of the response factor for the analyte. Additionally, a ten-fold dilution was employed in analyzing TBP in high hydroxide and evaporator salt solutions. No dilution was found necessary for the low hydroxide salt solution. These techniques were checked against prepared standards for the salt solutions being tested. The recovery efficiencies obtained using these methods ranged from 81% for high hydroxide salt solution to 89% for evaporator salt solution to 95% for low hydroxide salt solution.⁷

An initial analytical method was developed for this testing. During the early part of this experimental program, this method was found not to be as accurate as desired. As a consequence, results from some the first analyzed samples are not considered to be as accurate as later analysis. This is particularly believed to be the case for the earlier 40°C tests without any sludge.

The measurement of n-butanol from aqueous salt solutions requires separation of the n-butanol from the salt solution matrix. The currently accepted method for making this analysis involves conventional purge and trap preconcentration followed by gas chromatography. Both matrix and sparging geometry effects make this method both imprecise and inaccurate. Analysis of prepared standard solutions using this method, gave results that were 17 – 20% high.

An isotope dilution method was used to enhance the existing Gas Chromatography – Mass Spectrometry (GC-MS) methods for the determination of n-butanol for the last two data points of the 40°C kinetics test and for solubility, equilibrium and evaporator tests that followed. In the isotope dilution method, a known fixed amount of an isotopically labeled form of the compound of interest (in this case a deuterated form of n-butanol) is added to all samples. The deuterated n-butanol acts as an internal standard and effectively normalizes all matrix and sparge geometry effects. In addition to isotope dilution, the mass spectrometer was programmed for the selective ion monitoring mode. This reduction in region to be monitored helps to improve signal to noise ratios and improve sensitivity.

Tri-n-butyl phosphate hydrolyzes in the presence of strong alkaline solutions.¹⁰ Products of the reaction are dibutyl phosphate and 1-butanol. Further hydrolysis of DBP results in the formation of monobutyl phosphate and 1-butanol. Hydrolysis of MBP yields phosphate and 1-butanol as reaction products. These reactions are shown below in equations 1a – 1d. DBP is reported to be considerably more resistant than TBP to alkaline hydrolysis.¹⁰

If it is assumed that hydrolysis takes place only in the aqueous phase, when the overall TBP concentration is greater than the solubility of TBP in the solution as indicated in the literature,¹⁰ equation (2) becomes

TBP Hydrolysis Kinetics Experimental

Previously, Swingle¹¹ examined the hydrolysis of tri-n-butyl phosphate in simulated In-Tank Precipitation (ITP) salt solutions and observed slow kinetic decomposition. The tests examined relatively low temperatures (30 and 40°C). In the current tests, the effects of higher temperature, hydroxide concentration, and sludge concentration (catalytic decomposition) on the degradation of tri-n-butyl phosphate were investigated. A statistical design was developed to determine the effects of temperature, salt solution composition and sludge concentration, while minimizing the number of tests (Table 1). This design allowed determination of the primary effects of each variable. In the tests salt solution was added to glass vessels. Due to the low solubility of tri-n-butyl phosphate in these aqueous solutions, the concentration of the TBP required a destructive analytical approach. Therefore, these tests were run in a number of glass serum vials. Figure 1 shows some of the glass vials in the oven. A known quantity of tri-n-butyl phosphate (exceeding the solubility by more an order of magnitude) was added to each vial containing the salt solution and sludge (if required). The vials were placed in pre-heated ovens. Periodically through the test, vials were removed from the oven and analyzed for total TBP, DBP and butanol. For TBP, this included both soluble and insoluble fractions. Table 1 gives the initial statistically designed test matrix examining the effects of temperature from 40°C to 110°C, free hydroxide concentration from 0.21 M to 6.8 M and the presence or lack of sludge.

Butanol Production

The results of the TBP hydrolysis kinetics tests are presented in Figures 2 – 14. In general, all tests show decreasing TBP concentrations and increasing DBP and butanol concentrations with time as is expected. Table 3 gives the average ratios of concentrations of species of interest. The raw butanol production was found to be higher than expected, while the raw DBP production was found to be lower than expected. After accounting for the analytical recoveries of these two species, the butanol production for most of the low hydroxide and evaporator salt solution tests more closely matches the predictions for the production of one butanol atom from each atom of TBP hydrolyzed. The DBP production closely matches predictions in three of the four low hydroxide salt solution tests while the evaporator test DBP results improve significantly.

The high butanol and low DBP productions could indicate further decomposition beyond DBP. However, if that were the case, the ratio of the butanol to DBP concentrations would be expected to be constantly increasing. In most tests Figures 2 – 14 do not appear to show this. Instead it appears that unexpected concentration ratios are more likely the result of experimental and analytical uncertainty.

* These tests did not deplete the TBP. Therefore, the ratio of DBP/TBP or BuOH/TBP could not be calculated.
** DBP data not available.

Figure 2 presents the results of the test run in low hydroxide salt solution with no sludge at 40°C. Figure 3 gives the results of the test run in low hydroxide salt solution with 2,000 mg/L sludge at 40°C. The results for the test run in evaporator salt solution with no sludge at 40°C are given in Figure 4. Figure 5 shows the sample results from the test run in evaporator salt solution with 2,000 mg/L sludge at 40°C.

The results of analyses for TBP in the early part of both of the 40°C tests with no sludge are considerably below what was expected and what was initially added to the vials. This is believed to be because these samples were analyzed with an early version of the analytical method that was not as accurate as needed. Because these two initial samples were well below what was known to have been added, the initial TBP concentrations for these samples was taken from the known amount added. Later samples in those tests, as well as all results from the tests in which sludge was included appear to be within expected ranges. These samples were analyzed after the TBP analytical method was more completely developed.

The 40°C low hydroxide salt solutions show increasing concentrations of DBP and butanol with decreasing TBP concentration. This is expected with the hydrolysis of TBP to DBP and butanol. The ratio of the DBP and butanol concentrations seems also to be fairly constant as would be expected if there were no further decomposition of the DBP to MBP and butanol. The 40°C tests in evaporator salt solution show initially increasing DBP and butanol concentrations with a later drop in the concentrations of both species. This might indicate further decomposition of DBP to MBP and butanol, but a continuously increasing butanol concentration would be expected in that case, along with a DBP concentration that peaks and then begins to decrease. In no case are both of these phenomena apparent. So it is unclear what is happening with DBP and butanol in the 40°C evaporator salt solution tests, but it would appear to be experimental or analytical uncertainty.

Figures 15 – 27 show the TBP concentration data only along with concentrations predicted by the model.

At 40°C, both tests that were run without the presence of sludge resulted in lower than expected initial concentrations of TBP. This is believed to be the result of an incompletely developed analytical method at the time. Later samples from the same tests, as well as the samples from other tests, were analyzed using a more completely developed method. Though there are some differences between the hydrolysis rates at 40°C, the uncertainty makes them statistically the same. Therefore, there do not appear to be any effects caused by differences in salt solution composition or the addition of sludge.

SRTC performed tests to determine the solubility of TBP, DBP, and n-butanol in water. The purpose of determining the TBP solubility is to help determine the hydrolysis rate. The purpose of determining the DBP solubility is to determine whether it will form a floating organic phase or follow the salt in high level waste solutions. The purpose of determining the n-butanol solubility is to determine its tendency to form a floating organic layer in a waste tank.

Two methods were employed to measure the species solubility. In the first method, solutions of low hydroxide, high hydroxide, or evaporator simulants were placed in 500 ml (DBP, n-butanol) or 1 liter (TBP) flasks. TBP, DBP, or n-butanol were added to the flasks until a separate floating organic phase was observed. The samples were placed in an oven for several days. Periodically the flasks were inspected. If the separate floating organic phase was not observed, additional material was added. Periodically, samples were collected and submitted for analysis. The samples were collected from the bottom of the vessels to ensure only soluble TBP, DBP, and n-butanol was measured.

The analytical method was not able to measure the TBP concentration accurately enough to determine the solubility by this method. Since n-butanol is volatile and the tests were conducted at 40, 70, and 110°C, most of the n-butanol evaporated and was not measured. The DBP results were very inconsistent.

Therefore, a second method was used to measure the TBP, DBP, and n-butanol solubility. Solutions of low hydroxide, high hydroxide, and evaporator simulants were placed in sealed 60-mL serum vials. The solutions were heated to the desired temperature. TBP, DBP, or n-butanol were injected with a syringe into the vials until an insoluble organic phase was observed. If an organic phase was observed, the sample remained in the oven for an additional time and was periodically checked. The syringe was weighed before and after species addition to determine the amount added.

Table 5 shows the measured solubilities. Because of its low solubility (~ 1 – 2 mg/L), the TBP solubility could not be measured by this method. The TBP solubility in concentrated salt solution has been measured previously at ambient temperature (4.8 M Na, 1.5 M OH) and found to be 1.1 mg/L.¹⁴ The literature shows TBP solubility decreases with increasing temperature.¹⁰ The TBP solubility (1.1 mg/L) can be considered bounding for high hydroxide and evaporator solutions at ambient temperature and higher temperatures. The TBP hydrolysis rates were measured in another part of this work.

The DBP solubility was very high in high hydroxide (50 – 58 g/L) and low hydroxide (15 – 19 g/L) simulants. The variations observed with temperature are probably due to experimental uncertainty. The results indicate DBP will be very soluble in pump tanks and waste tanks, and should be transported with the salt solution rather than accumulate as a floating organic phase. The DBP solubility in evaporator simulant was much less than in the high and low hydroxide solutions. These results suggest there is a hydroxide effect and a sodium effect on the DBP solubility. Increasing the hydroxide concentration appears to increase the DBP solubility, while increasing the sodium concentration appears to decrease it.

The measured n-butanol solubility in high and low hydroxide simulants at 40°C agrees well with previous SRTC work.¹⁵ Previous SRTC work measured an n-butanol solubility of 5.8 g/L at 35°C in a 4.8 M Na standard salt solution (1.48 M OH) and an n-butanol solubility of 3.6 g/L in a 6.0 M Na high hydroxide salt solution at 35°C. The results show the n-butanol solubility decreases with increasing temperature. The technical literature shows the solubility of n-butanol in water decreases with increasing temperature.¹⁶ Previous SRTC work measuring the solubility of n-butanol in salt solution found the solubility is not significantly affected by temperature.¹⁵ The solubility decreases significantly in the evaporator solution. The results show butanol has a high solubility in the high and low hydroxide solutions. Since butanol is very volatile and has a high solubility, the formation of a floating layer in pump tanks or waste tanks is unlikely.

The conclusions of the solubility tests are the following: TPB solubility in High OH, Low OH, and evaporator solution is very low. DBP will be very soluble in pump tanks and waste tanks, and should be transported with the salt solution rather than accumulate as a floating organic phase. The formation of a floating butanol layer in pump tanks or waste tanks is unlikely.

The solubility of butanol is significantly lower in evaporator salt solution than in either low or high hydroxide salt solution. This phenomenon was also noted with DBP and is probably due to high sodium concentration or a "salting out effect." If salt solution containing butanol in concentrations greater than the solubility limit for the evaporator reaches the evaporator, the butanol would be expected to flash to the vapor phase because of the reduced solubility and high temperature.

Since the TBP decomposition process occurs only in the aqueous phase, the process can be divided into two steps: transport of TBP from a floating organic layer into the aqueous phase and hydrolysis of the TBP molecules in the aqueous phase. The decomposition rate will be controlled by the rates of the two processes.

If the mass transport rate is much slower than the hydrolysis rate, the aqueous phase will have a low concentration of TBP. As soon as the TBP is transported to the aqueous phase, it will decompose and the decomposition rate will be controlled by the mass transfer rate.

If the TBP hydrolysis rate is much slower than the mass transport rate, the aqueous phase will be nearly saturated with TBP. The high TBP concentration will reduce the rate of transport of TBP from a floating organic to the aqueous phase and the TBP decomposition rate will be controlled by the hydrolysis rate.

SRTC performed tests to measure the transport rate of TBP from a floating organic layer into aqueous solutions of de-ionized water and 1M NaOH. The purpose of the tests was to determine the TBP diffusivity, which can be employed to calculate mass transport rates in High Level Waste Pump Tanks and Waste Tanks. The mass transport rates can be compared with hydrolysis rates to determine the rate-limiting step and to calculate TBP decomposition rates in tanks.

SRTC performed the mass transfer tests in the following manner. The SRTC Glass Shop fabricated a glass vessel with sample ports located 1, 2, 3, 4, 5, and 6 inches from the bottom (see Figure 23). The vessel was filled with de-ionized water or 1M NaOH to the 6.5-inch level. Tri-n-butyl phosphate was added to the vessel until a floating layer covered the surface. Samples of the aqueous phase were collected at different levels and submitted to ADS for TBP analysis. The tests were performed at ambient conditions. De-ionized water and 1 M NaOH were used in this test rather than the simulants discussed previously because of the large sample volume needed to measure TBP with those simulants.

Mass transfer tests were run according to the description above. The researchers observed a large variability in diffusion coefficients calculated from the data. Additionally, no significant concentration gradient was observed. The results suggest thermal convection was a factor in the TBP mass transfer.

Because of the large uncertainty in the mass transfer tests and the presence of thermal diffusion, SRTC decided to use diffusivities for TBP in water found in the technical literature to evaluate the relationship between mass transfer and hydrolysis in High Level Waste Pump Tanks and Waste Tanks. The diffusivity of TBP in water at 25°C is 2.29 x 10^-6 cm2/sec.⁷ With this data, one can calculate the diffusivity at other temperatures using the Stokes-Einstein, Wilke-Chang, or Tyn-Calas equations which are described by equation [6]

Since thermal gradients will exist in waste tanks, thermal diffusion must also be considered. The thermal gradients will cause TBP density variations that will increase the mass transfer rate. The thermal convection mass transfer coefficient can be calculated by using an analogy with thermal convective heat transfer that is described by equation [7].¹⁸

Table 6 shows the parameters and calculated mass transfer coefficients at 25°C, 40°C, 70°C, and
110°C.

These results show that in an unagitated tank at a temperature of 25°C the mass transfer and hydrolysis rates are the same order of magnitude. However, the mass transfer resistance in an unagitated waste or pump tank at 25°C causes a reduction in the butanol generation rate below that which would be predicted by the kinetics of TBP hydrolysis alone. At temperatures of 40 to 110°C in an unagitated tank, the mass transfer rate is 1-2 orders of magnitude slower than the hydrolysis rate. Under these conditions, the actual butanol generation rate would be less than predicted by the hydrolysis test data. The conditions in Table 6 provide an example of the relative influence of mass transfer and hydrolysis in the TBP decomposition process. The diffusivity measured in water and the parameters calculated for 1 M NaOH are expected to bound the diffusivity and mass transfer rate in High Level Waste salt solution. However, since the analysis required extrapolation of the data, caution should be used in applying this data to High Level Waste Tanks.

If a floating organic layer is observed in a waste tank or pump tank, this approach could help High Level Waste determine the n-butanol generation rate.

Samples of high hydroxide salt solution containing ~100 mg/L TBP were prepared and sealed in serum vials similar to those used in the TBP hydrolysis tests. Samples were irradiated to 28.75, 53.75 and 100 Mrad in SRTC’s Shepherd ⁶⁰Co Source. The temperature inside the ⁶⁰Co Source at the time of the irradiation tests was 37°C, and the irradiation took place over several days. The samples were then analyzed for TBP, DBP and butanol using the analytical methods described above. The 100 Mrad exposure corresponds to a number of years in the tank, depending on the tank.

TBP, DBP and butanol results for the TBP irradiation tests are given in Table 7. The TBP is decomposed over time due to irradiation though apparently much slower than expected from hydrolysis. Essentially no butanol nor any other volatile or semivolatile organic compound was found in these samples. It is not clear what happened to the TBP in these tests, but in general radiolysis of organic materials leads to formation of shorter chained soluble non-volatile organic acids. These acids are not detectable by the methods used to analyze these samples.

In the evaporator test, an evaporator salt solution containing 130 mg/L of butanol was placed in an evaporator pot (see Figure 24). There was a water-cooled condensing coil to condense water and butanol from the vapor phase. The temperature of the salt solution was increased until it reached the boiling point (~116 °C). The system was allowed to equilibrate at that temperature for about an hour. At that time a sample of the vapor phase was pulled along with condensate sample. The salt solution was allowed to cool slowly and vapor phase samples were pulled at 110, 100, 90, 80 and 60°C. Only a small sample of condensate was obtained at 110°C and none at any temperatures below 110°C.

One potential scenario in the Tank Farm is an accumulation of a TBP/n-paraffin layer on top of the aqueous layer in which butanol may accumulate. Distribution of butanol between aqueous, organic and liquid phases was examined in a series of equilibrium tests. Dr. Harry Babad, who was hired as a consultant on this program, recommended these tests.¹⁹ In these tests, high hydroxide salt solution containing varying amounts of butanol was placed in a test vessel with or without a mixture of 30% TBP in n-paraffin. The mixture was allowed to equilibrate at a constant temperature. Then samples of both the vapor and aqueous phases were taken and analyzed for butanol using gas chromatography. Figure 25 shows a test vessel being sampled. Table 8 gives the test matrix for the equilibrium tests. This series of tests were hoped to measure the Henry’s Law constant for butanol for each system and determine the distribution of butanol in the system.

Analyses for organic materials in aqueous and surface floating samples taken from the 5-H Pump Tank and Waste Tank 38H and in vapor samples taken from the 5-H and 3-F Pump Tanks have been completed. The results indicate that the concentration of organic materials is extremely low in all samples.

Some organic materials were found in the HPT-5 and FPT-3 vapor samples but in nanogram/liter (ng/L) quantities. These materials were present in the samples at levels a factor of ten below the practical quantitation limit. Because of the low levels and the fact that no field background analysis was run (laboratory background analyses were run), the researchers cannot absolutely determine whether the materials were actually taken from the pump tanks or whether they are from environmental background. In any case, the quantities of material found are several orders magnitude below that which would comprise a flammability concern.

Small quantities of organic materials were also found in the surface floating samples taken from Tank 38H and the 5-H Pump Tank. The Tank 38H sample contained small quantities of siloxane-type materials. Though it is possible that these materials result from decomposition of the solid phase extraction disk used to pull the sample, the composition of the material indicates that it is more likely composed of decomposition products of one of the siloxane-based defoamers used in the high level waste handling process. The 5-H Pump Tank sample contained small amounts of toluene and other materials that may have come from scintillation cocktail material used in radiochemical analysis and that probably originated in the DWPF lab. The 5-H Pump Tank sample also contained small amounts of material that may have resulted from decomposition of ion exchange resins (several substituted benzene ring compound and one styrene compound). All of these materials were present in small quantities (hundreds of m g) on the sample disk. Because of the sampling method, converting the quantities measured to a meaningful concentration in the tanks is not possible, but it appears that all materials measured are present in very small quantities in the tanks. Further details on the analyses and results may be found in the report documenting the initial sampling effort.²⁰

With the exception of small amounts of some organosilicon compounds, no quantifiable organics were found in floating organic samples taken from either the Tank 26F or Tank 33F. These organosilicon compounds may actually be present in the tanks (e.g., from organosilicon-based lubricants or antifoaming agents) or may be artifacts of the solid phase extraction (SPE) disks used in sampling. Though quantification of the concentrations of the compounds in the tanks is not possible because of the sampling method, the results indicate that the concentrations are low.²¹

A small amount of amount of tri-n-butyl phosphate (TBP – 1.3 mg/L) was found in the dip sample taken from FPT-3. A somewhat larger amount of TBP (230 mg/L) along with some normal paraffin hydrocarbons (NPH – 77 mg/L) were found in the variable depth sample from the pump tank. No other detectable volatile or semivolatile organics were found. Small amounts of three different organosilicon compounds were found on the floating sample solid phase extraction (SPE) disk as well as a trace of xylenes (3m g/disk) and a trace of ammonia (120m g/disk). No other volatile or semivolatile organic compound was detected in the FPT-3 samples.²²

No quantifiable organics were found in the either the Tank 26F dip (surface) or variable depth sample (VDS - subsurface) samples or the Tank 46F dip sample. The total organic carbon analysis for the two Tank 26F samples indicated the presence of measurable amounts of organic carbon, but no quantifiable individual species were found. Neither semivolatile or total organic carbon analysis could be run on the Tank 46F sample.²³

The only quantifiable organic species found in the Tank 33F dip and VDS samples were non-volatile formate ion and in the subsurface VDS sample a small amount of normal paraffin hydrocarbon (NPH).

No quantifiable organic species were found in either the dip or VDS samples taken from Tank 43H, though total organic carbon analysis did indicate the presence of measurable organic carbon. Analysis of a floating organic sample taken from Tank 43H indicated the presence of small quantities of a number of volatile and semivolatile organic compounds. Because of the sampling method, converting quantities measured to a meaningful concentration in the tanks is not possible, but it appears that all materials are present in very small quantities in the tanks.

Some organic materials were found in the vapor samples taken from High Level Waste Tanks 26F, 33F, 46F, 11H, 22H, 32H, 39H and 43H, but only in nanogram/liter (ng/L) quantities. The quantities of material found are several orders magnitude below that which would comprise a flammability concern.²⁴

In some instances, the organic species may form a floating layer that is immiscible with the aqueous salt solution. High Level Waste asked SRTC to determine the auto-ignition temperatures of organic species in the Tank Farm to determine the plausibility of the sun guns igniting any floating organic layer that formed in waste tanks. In addition to the auto-ignition temperatures, HLW asked SRTC to determine whether a spark could ignite a floating organic layer in the tank.

High Level Waste provided SRTC with a list of organic species in the Tank Farm. SRTC examined the National Fire Protection Association (NFPA) codes and the Material Safety Data Sheets (MSDSs) to determine the auto-ignition temperatures for these species. The lowest auto-ignition temperature identified was 190°C for trimethylamine.

The analysis showed ignition of a floating organic layer in a high level waste pump tank or waste tank is not a credible event. Details of the analysis can be found in a previous SRTC memo.²⁵

High Level Waste Engineering requested SRTC to evaluate agitators for mixing sludge and organic layers in pump tanks. SRTC performed the evaluation by reviewing previous SRTC Canyon mixing studies, reviewing the AEA Pulsed Tube Mixer test report, and conducting a literature search. The analysis found the AEA pulse tube mixer should be able to suspend and mix sludge in the pump tanks, but has not demonstrated its ability to disperse a floating organic into an aqueous liquid. Testing is needed to determine whether the pulse tube mixer can disperse floating organic layers into aqueous liquids. In theory, an agitator could be used to disperse floating organic layers into aqueous liquids. A variety of impeller types could be used including flat blade impellers and pitched blade impellers. Typical impeller speeds would be approximately 200 – 300 rpm. However, installing the agitator and performing needed maintenance could present obstacles that are difficult to overcome. In addition, some testing would be advisable prior to design and installation of an agitator system. Details of the analysis can be found in a previous SRTC memo.²⁶

The results of the tests documented in this report suggest opportunities for future work to reduce uncertainties in these studies. Possible items to be addressed are given below.

• Improve analysis for TBP to reduce the limit of detection to allow better measurement of solubility.
• Develop analytical method for monobutyl phosphate (MBP) to allow better study of kinetics of both DBP and MBP hydrolysis.
• Conduct tests to better understand the possible effects of organic layer contact area on TBP hydrolysis rate.
• Consider new approaches to the butanol equilibrium tests. One possible approach would be measuring vapor phase butanol content by measuring pressure change in a sealed vessel.
• Consider new approaches to determine whether results of evaporator test represent real phenomena or are an artifact of the experimental method.

The customer requirements were specified in HLE-TTR-98-065.

Quality assurance measures were directed by task technical and quality assurance plans for this program (WSRC-RP-98-01261, Rev. 1).^4,5 Among the quality assurance measures specified in the task plans was periodic checking of the oven temperature with M&TE thermometers traceable to NIST standards. Inadvertently, these checks were not completed during the initial set of tests. Upon discovery of this oversight a Program Deficiency Report was initiated and compensatory actions were taken.²⁷ The oven in 679-T used to complete the solubility tests and the 70 and 110°C TBP hydrolysis tests was checked after completion of the tests with a calibrated thermometer, and the digital read-out on the oven was found to agree within 2°C at 70° and 110°C and within 5°C at 40°C. Calibration of the 679-T oven was found not to be readily adjusted. The oven in 773-A used for the 40°C tests and the butanol equilibrium tests could not be checked, since its calibration was easily adjusted and indeed found to have been changed. Comparison of the data from the 40°C hydrolysis tests with the data from the 70 and 110°C tests using an Arrhenius plot (Figure 22) indicates that the 40°C hydrolysis test was likely run at 40°C. (The data for the all three tests define a straight line as would be expected for a reaction that is first order in TBP concentration.) During the tests performed at 150°C, 30°C, and 70°C (with dilute salt solution), the oven temperature was checked with a calibrated thermometer during the test.

Data for this experimental program are recorded in laboratory notebooks WSRC-NB-99-203 and WSRC-NB-99-00097.