ML20134M190

Requests to Discharge Unit 2 Waste Neutralization Sump

ML20134M190
Person / Time
Site:	Millstone
Issue date:	02/10/1997
From:	Scace S NORTHEAST NUCLEAR ENERGY CO., NORTHEAST UTILITIES SERVICE CO.
To:	Grier J CONNECTICUT, STATE OF
References
SES-97-GN-021, SES-97-GN-21, NUDOCS 9702200135
Download: ML20134M190 (9)

Text

_..- _ - _

m_

4

~

Nordmast msms=,aerosat Untities System mtm sm%

P.O. Bos 270 Hented,cr ost41.orro (sos)e65.sooo February 10,1997 l

- SES-97-GN-021 l

310802 Mr. James Grier Supervising Sanitary Engineer l

Water Management Bureau i

Department of Environmental Protection 79 Elm Street Hartford, CT 06106-5127

Dear Mr. Grier:

i Millstone Station i

NPDES Permit No. 0003263 Reauest to Discharne Unit 2 Waste Neutralization Sumo I

I As previously discussed with you in your office February 7,1997, Northeast Nuclear Energy Company (NNECO) is seeking confirmation that it may discharge wastewaters from the Condensate Polishing Facility (CPF) Waste Neutralization Sumps at Unit 2 through DSN 001B-6 so long as such wastewaters contain no hydrazine above detection limits.

At present, even though Unit 2 is not producing power, wastewaters are continuing to accumulate in the Unit 2 CPF Waste Neutralization Sumps. One of the two sumps, TK-10, is currently at capacity (approximately 25,000 gallons) and the second sump, TK 11, is approximately at 40% capacity. While there are no on-going processes during shut-down which would add significant waste water into these sumps, minor wastewater amounts are being added daily, primarily from the heating boiler exhaust steam drain which contains no hydrazine. Nevertheless, spectrophometric analysis of a sample recently taken from the TK 10 indicates hydrazine in small amounts (282 ppb).

As discussed with you on January 28,1997, NNECO has performed laboratory evaluations of the applicability of spectrophotometry for detecting hydrazine in waste waters with high total suspended solids (TSS)c It is our conclusion that this method gives false positives for hydrazine. Results of these investigations, which.were undertaken in-support of a separate request (to be submitted under separate cover) to establish a de minimus concentration level for hydrazine, are included here as well (Attachment 1).

I

\\

hf'g 9702200135 970210 s

PDR ADOCK 05000336' p

PDR ;

190083

However, separate and apart from any de minimus determination, in that the CPF sumps i

are currently at or near capacity, we have performed additional hydrazine analyses in NNECO's laboratory using spectrophometry on a sample from TK 10. These analyses i

were undertaken both before and after extracting TSS using a coagulant in conjuction with sequential filtration (Attachment 2).

I Once the TSS interference was removed, our tests indicated hydrazine at 2.04 ppb which is less than the detectable level for spectrophotometry. Experimental precautions were i

taken to ensure that the coagulant and filters were not themselves removing hydrazine from the samples. As indicated in Attachment 2, the method employed was extemely time intensive taking up to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. It would not,' therefore, be suitable or practicle for routine j

spectrophometric analysis of hydrazine.

i Accordingly, NNECO requests a determination that wastewater contained in the Unit 2

)

Condensate Polisher Waste Neutralization sumps can be discharged via DSN-001B-6 so j

long as hydrazine is not found above detectable levels using spectrophotometry when adjusted for TSS interference.

Should you have any questions, please call Mr. Paul Jacobson, Emironmental Services -

Nuclear at (860) 447-1791 Ext. 2335.

Very truly yours, j

NORTHEAST NUCLEAR ENERGY COMPANY Cnc G r

S.E. SCACE Director - Nuclear Engineering Programs cc: M. Harder hTC i

i l

\\

~

l i

i Evaluation of Spectrophotometric Analysis for Hydrazine i

This report summarizes results of an evaluation of the UVNIS spectrophotometric analysis i

method for hydrazine. Specifically, this report explores how this analysis performs under chemical matrix conditions expected in a typical waste neutralization sump or waste drain tank.

NNECO sought to determine which method to employ in streams where hydrazine may be present only at ppb levels. In this study, samples of both the waste neutralization sump and waste drain tanks were used in the laboratory evaluation.

I Waste Neutralization Sumo l

A sample was taken from the Unit 2 waste neutralization sump TK-10 (which discharges to DSN 001B-6) following an appropriate recirculation period. The water in this sump was present as a result of draining evolutions associated with lay up of the resins. While, this particular batch of water is not expected to contain hydrazine, the chemical matrix of the wastewater is i

representative of that which will be present during plant operation (high total dissolved solids

)

and corrosion products).

1 The sample was initially analyzed by iodine titration and found to contain less than the j

' minimum detectable level of 350 ppb hydrazine Spectrophotometric analysis of this same sample then yielded a result of 271 ppb hydrazine. The sample was then processed through a j

}

series of glass fiber filters, each with a diminishing particle size, and the collected filtrate was analyzed for turbidity as well as hydrazine. The results are listed in Table 1 and show a clear correlation between the amount of suspended matter in solution and the measured j

concentration of hydrazine.

i Table 1 TK-10 Analysis i

Sample Hydrazine Concentration Turbidity (NTU)

(ppb)

Raw 271 22.5 10 micron effluent 154 15 8 micron effluent 139 13 5 micron effluent 122 11 1

0.8 micron effluent 92 10 0.45 micron effluent 81 7.5 j

The relationship between turbidity and hydrazine in the TK-10 sample is graphically illustrated in Chart 1. Linear regression analysis was employed to construct a best fit curve of the laboratory data. The equation and correlation coefficient of the curve are included with the chart. The curve was then extrapolated down to 2.8 ppb to determine the minimum turbidity f

level necessary to generate a hydrazine concentration equal to the limit of detection (LOD) of i

the spectrophotometric method. Based on that analysis, turbidity is equal to approximately 2.5 i

nephelometric turbidity units (NTUs).

n

~

n

i Chart 1 TK-10, Turbidity vs Hydrazine Concentration Turbidity vs Hydrazine 30 25 -

y = 0.076ex + 2.3171 R' = 0.9788 I

20 -

15 -

5 10 -

g 5-0 l

0 50 100 150 200 250 300 Hydrazine Concentration (ppb)

Another sample of TK-10 was collected and processed in a similar fashion to further test the consistency of this effect. Aliquois of the filtrate were also spiked with a known quantity of

.hydrazine and the percent recovery was calculated. These results are listed in Table 2.

Table 2 TK-10 Analysis with Spike Recovery Sample Hydrazine Turbidity (NTU)

Hydrazine Percent Concentration Concentration after Recovery (ppb) 35 ppb spike (ppb)

(%)

Raw 273 22.5 304 88.6 10 micron effluent 153 15 183 85.7 8 micron effluent 128 13 160 91.4 0.8 micron effluent 88 10 129 117 0.45 micron effluent 86 7.5 108 62.9 The results listed in Table 2 indicate the spectrophotometer is capable of disceming relatively small changes in hydrazine concentration (as witnessed by the percent recovery values), yet it is quite susceptible to interference from matter suspended within the solution.

In an effort to gather more information, a sample of TK-11 was drawn and submitted to the same filtration and analysis techniques. The water in TK-11 is from the same source as that of TK-10. Even so, it was noted that the color of the suspended matter in the TK-11 sample was decidedly different from that seen in the TK-10 sample. In contrast to the dark black of the TK-10 material, the solid matter in the TK-11 sample was light brown. This may be due to a difference in the oxidation state of the corrosion products (which make up the bulk of the filterable solids), although there is no analytical data to support this hypothesis. The results from the TK-11 analyses are presented in Table 3.

2

Table 3 TK-11 Analysis with Spike Recovery Sample Hydrazine Turbidity (NTU)

Hydrazine Percent Concentration Concentration Recovery (ppb) after 35 ppb

(%)

spike (ppb)

Raw 203 25 193

-28.6 1

10 micron effluent 44 15 60 45.7 8 micron effluent 41 13 58 48.6

)

0.8 micron effluent 35 10 54 54.3

, 0.45 micron effluent 29 8

49 57.1 The effects from the suspended inatter on the measured hydrazine concentration in the TK-11 sample were greater than those encountered in the TK-10 sample. This is clearly evident in the-linear regression curve constructed for the TK-11 data (Chart 2). The most likely reason is the difference in color between the two samples, it would appear that interference of the spectrophotometric analysis is a function not only of the amount of suspended matter but also the oxidation state, making hydrazine quantification all the more difficult. Filtration will reduce J

the turbidity, but will be ineffective on color related interference.

Chart 2

)

TK-11, Turbidity vs Hydrazine Concentration Turbidity vs Hydrazine 1

35 y = 0.0837x + 8.3047 8

30 -

R = 0.887 20 -

7 if -

k 10 -

5-0 O

50 100 150 200 250 300 i

Hydrazine Concentration (ppb)

The extent of analyticalinterference was further pursued by performing a spectrographic scan of the TK-10 & 11 raw samples. The " scan" function of the Perkin Elmer Lambda 2 Spectrophotometer provides a measurement of absorbance versus wavelength for a given sample. At the 458 nm wavelength used for hydrazine analysis, TK-10 exhibited 0.75 absorbance units while TK-11 exhibited 0.90 absorbance units. Both samples had a similar 3

l.

j l.

mass of suspended matter. As a reference, a 100 ppb concentration of hydrazine in a pure l

water matrix will exhibit 0.128 absorbance units.

The manufacturer of the instrument was contacted to obtain further insight. Perkin Elmer provided a curve of absorbance versus wavelength derived from a turbidity method developed for the Lambda 2. The curve clearly shows that the highest absorbance occurs in the

' wavelength range of 360 to 600 nm. Since the hydrazine absorbance is measured at 458 nm, this infers that even modest levels of turbidity will produce a false positive indication of hydrazine. The Perkin Elmer representative reiterated that low level hydrazine analysis should f

only be performed on samples with a pure water matrix. According to Perkin-Elmer, samples f

with a turbidity of 1 NTU or greater will produce inaccurate results for hydrazine. As i

demonstrated above, sequential filtration down to 0.45 microns was unsuccessful at producing an acceptable turbidity in the waste neutralization sump samples, j

i An attempt was then made to compensate for turbidity effects by using an untreated (no reagents added) cuvette of the waste neutralization sump water as a blank. The blank solution.

had the same product mass and oxidation state as that of the solution being analyzed for i

j hydrazine. Repeated trials produced the same result: an error message. Once again, Perkin l

Elmer attributed this to the turbidity level of the solutions.

i A final attempt to remove the turbidity consisted of sample treatment with anion exchange resin. Another sample of TK-10 was collected and sequentially processed through glass fiber 3

filters of diminishing pore size, culminating at 0.22 microns. This served to reduce the turbidity from 25 NTU's to 3.5 NTU's. The collected filtrate was then passed through a column i

containing Dowex* monosphere anion resin. If the suspended matter causing the remnant l

turbidity existed as a negatively charged species, the anion resin would functionally remove it l

without affecting any hydrazine that may be present.

i j

The anion resin effluent was then analyzed for turbidity, with the results showing 3.5 NTU's; no i

' change. An aliquot of the anion effluent was then prepared for spectrophotometric analysis, along with a 35 ppb hydrazine controlled standard, prepared in a pure water matrix. After waiting the appropriate amount of time for color development, the control standard exhibited 1-the familiar yellow hue seen in hydrazine bearing samples. No color change was observed in the anion resin effluent sample. Subsequent analysis in the spectrophotometer produced a result of 34 ppb for the control standard and 28 ppb for TK-10.

Therefore, the fact that a sample with no visible color development can produce a measured result close in concentration to one with a known quantity of hydrazine further confirms the turbidity-related interference and illustrates the difficulty in obtaining an accurate hydrazine value with the spectrophotometric method on waste streams containing measurable turbidity.

o Significantly, the presence of suspended solids in solution had little, if any, effect on the iodine titration method of analysis. Samples from TK-10 & 11, utilized with the spectrophotometric experiment, were spiked with known quantities of hydrazine and analyzed by iodine titration.

The results are presented in Table 4.

l i

4 4

Table 4 TK-10 & 11 Spike Recovery with lodine Titration Sample TK-10 Percent TK-11 Percent Description Recovery (%)

Recovery (%)

Raw

< 0.35 ppm N/A _

< 0.35 ppm N/A 10 ppm Spike 11 ppm 110 12 ppm 120

_ 25 ppm Spike 26 ppm 104 27 ppm 108 50 ppm Spike 54 ppm 108 51 ppm 102 100 ppm Spike 97 ppm 97 103 ppm 103 Waste Drain Tank A testing regimen similar to that employed with the waste neutralization sump samples was used with the aerated waste drain tank water which discharges via DSN 001B-2. A sample from the "A" aerated waste drain tank was analyzed for hydrazine using the iodine titration method and found to contain less than the minimum quantifiable level of 350 ppb.

Spectrophotometric analysis of this same sample yielded a result of 42 ppb hydrazine. The sample was then processed through a series of filters and analyzed. Spike recovery testing was also performed. Turbidity analyses were unable to be performed due to a malfunction of the instrument. A summary of the results can be found in Table 5.

Table 5 "A" Aerated Waste Drain Tank Sample Hydrazine Hydrazine Percent Recovery Concentration (r'no)

Concentration after

(%)

30 opb spike (ppb)

Raw 42 53 36.7 10 micron effluent 12 25 43.3 0.8 micron effluent 9

22 43.3 0.45 micron effluent 8

22 46.7 The pattem of decreasing hydrazine for decreasing suspended mass was noted cbng with the poor percent of recovery. It was also noted that the color of the suspended matter in the aerated waste drain tank sample was similar to that observed in the TK-11 sample. Thus, it appears that turbidity and sample color effects are not confined to the waste neutralization sumps alone.

Summary Based on the hbove investigations, hydrazine analysis by UVNIS spectrophotometry is highly depend 3nt oa the quantity, as well as color, of undissolved matter in the sample being i

analyzed. Secuentital filtration is not a suitable remedy as repeated attempts in the laboratory were unsuccesful at reducing ' 'e turbidity to a point where interference was less than the LOD of the analy9A Furthermt

, filtration will not affect sample color. The use of spectrophotometry N hydrazine analysis should be restricted to samples without a measurable total suspended solids (TSS). This effectively eliminates analysis of the 5

i condensate polishing waste neutralization sumps by thit methodology and, unless the criteria of turbidity and color are reached by some processing regimen, the aerated waste drain tanks as well. As a result, the most appropriate and accurate analytical method for these streams is iodine titration.

t i

i T

n i

4 i

r I

1 l

l l

i 6

l l

r j

I, -

j Removal of Total Suspended Solid interferences From TK-10 Wasterwater Sample j

i l

1 Rec' nt work in the Unit 2 laboratory using a polyelectrolytic coagulant (Nalcolyte 7701), in l

e conjunction with sequential filtration, has served to remove turbidity (and related interference) l from a sample of TK-10 waste water. Subsequent analysis of the low-turbidity sample by j

UVNIS spectrophotometry poduced a hydrazine result that measured less than the minimum detectable concentration. A summary of the analyses results can be found in the following Table.

l TK-10 Analysis l

i Sample Turbidity (NTU)

Hydrazine (ppb)

Amount of Coagulant Identification Added (mLs)

)

Raw 23 282 0.20 I

10 micron effluent 8

75 0.45 micron effluent 3.5 12.7 i

0.22 micron effluent 2.5

[

0.22 micron effluent 1.5 l

0.22 micron effluent 1.0 4.5 l

0.22 micron effluent 0.84 0.10 l

0.22 micron affluent 0.72 2.66*

l 0.22 micron effluent 0.55 2.45*

{

0.22 micron effluent 0.45 2.25*

j 0.22 micron effluent 0.41 2.04*

1

• Result reported as 5 2.8 ppb j

t j

The familiar turbidity-hydrazine relationship, documented in the previous laboratory study i

(Attachment 1), was once again evident. By agglomerating the finely divided suspended solids i

into a more readily filterable mass, the coagulant used assisted in reducing the interfering turbidity to a point where it's effect on the spectrophotometer was minimal. The coagulant j

selected would not affect any hydrazine that may have been present in solution.

i it is worth noting the amount of time required to achieve turbidity reduction by this process, from sample acquisition to attainment of the minimum detectable concentration, was approximately 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. The requisite reaction period between the coagulant and the suspended matter, as well as the laborious, multiple filtration evolutions required to complete this procedure, make this type of sample method impractical for general use. Thus, hydrazine analysis by iodine titration remains the preferred method.

,