ML061000468

From kanterella
Jump to navigation Jump to search
E-mail from S. Sandike of Entergy to J. Noggle of USNRC, Regarding Latest from Quinn
ML061000468
Person / Time
Site: Indian Point  Entergy icon.png
Issue date: 11/22/2005
From: Sandike S
Entergy Nuclear Operations
To: Noggle J
Division of Reactor Safety I
References
FOIA/PA-2006-0081
Download: ML061000468 (19)
v • d • e


Text

t Jam'es -Nc,-aa'Ie -- L'ate's.:fro m-Q-uin-n Page 11 From: "Sandi<e, Steven" <SSandik@entergy.com>

To: <JC(NJnrc.gov>

Date: 11/22/05 10:15AM

Subject:

Latest from Quinn Stuff we worked on last night....

<

> <<NCD-SFDS-00-03.xls>> <<U2H3-2005-sources.xIs>> And some 80-10 data, along with the gamma spec cover sheet from the local well digging at U2 FSB wall <<80-1 0-etc.pdf>> I am still collecting and preparing a spreadsheet for other 80-10 data back to 2000...Although we have collected H-3 on Unit 2 storm drains for years, it has only recently been added to WinCDMS, so I am digging it out of the Tritium data sheets. (I will ariange for it to be back-fitted into WinCDMS as well - we need to see and manage this trend....). Steve Sandike Effluents /RMS.  :'!:.', - ,--O
' '- P fl4 -
. -r .: ENN Indian Point Energy Center Buchanan, NY 10511-0308 phone: )114-736-8455 fax: E1 4-734-6010 email: ssandik~entergy.com Date: November21, 2005 To: Dennis Quinn From: Lawrence Dauer

Subject:

Review of "Unit 2 Spent Fuel Pool Total Volume Loss Estimation" October27, 2005 Indian Point Unit 2 Spent Fuel Pool Investigation Team PURPOSE Review the volume loss estimate methodology and conclusions as documented in the uUnit 2 Spent Fuel Pool Total Volume Loss Estimation" memo dated October 27, 2005 (SEe Appendix A). The review will include the generation of regression results from the data, the determination of confidence intervals about the predicted slope, estimating the change in volume based on boron concentration, and a test of the utility of the model.

REVIEW Regression Data To check on the regression method, the boron concentration measurement data as listed on Attachment 1 of the October 27, 2005, volume loss estimate (Appendix A) were entered into an Excel spreadsheet. Linear regression analyses were performed usilg the Excel analyses tools (ie. Regression). In addition, other statistical analyses were performed.

The data is shown on Graph 1 attached and shows some variability over time that may represent sampling and analysis variability as well as the potential for cyclical changes as evaporation and water additions occur over time. The resulting regression data is summarized below:

  • S.- -- - .6 0

Boron Concentration (ppm) I Y Days since Cycle 17 Start X N (number of observations) 47 DF (degrees of freedom = N-2) 45 R__ 0.01 b [y-intercept] (ppm) 2407 m [slope] (Appm) -0.0202 SEm [standard error of slope] 0.0288 1 of 5

Confidence Interval About Slope When using regression analyses for estimating the value of a slope or change in a parameter over time, it is important to consider that statistical linear regression is typically performed using a least-squares fit methodology to the data. This is the case in boih the Excel Regression tool results above and in the volume estimate being reviewed. Therefore, the predicted (or estimated) slope can be considered the 'best fit' slope to the data, however the true slope will actually fall within a range of values. In order to assess this range, standard statistical regression analyses should utilize a confidence interval about the slope method. The method uses the tabulated critical t-value (tri) from the t-distribution, based on the degrees of freedom (DF) and a two-tailed confidence level of 95% (a= 0.05).

With a DF=45 for this data, the t~d(DF=45, a= 0.05) = 2.014.

The confidence interval about the slope is calculated from:

m +/- (tcix SEm)

In this case, the proper way to report the slope is with these confidence intervals. For the data in question, the estimated slope should therefore be listed as follows:

-0.0202 +/- 0.05808 ppm therefore, the true slope is expected to fall within the following range:

-0.0783 to 0.0379 ppm Estimating the Change in Volume Based on Boron Concentration The change in volume can be estimated from the estimated slope range, using an estimate of the total U2 SFP volume. As assumed in the volume estimate being reviewed, the volume of the Unit 2 SFP is based on dimensions of (30.67' x 34' x 39')

and does not take into account any reduction in volume taken up by spent fuel and fuel racks (note that because of this, the value may result in a slight overestimate in any change in volume). Total volume of the U2 SFP is estimated to be 304,200 gal. The estimated change in volume (gpd) is listed below:

3- *.~~ O- S

  • I. -

Lower 95% l -0.0783 l -3.253E-05 l -9.9 Calculated slope -0.0202 -8.393E-06 -2.6 Upper 95% 0.0379 1.57E-05 +4.8 2 of 5

Testing the Utility of the Regression Model for Predicting Changes The utility of the regression model for predicting changes in volume using the available data can be evaluated using a significance test of the regression slope. This is a standard statistical technique and is based on hypothesis testing. One can test the hypothesis if slope m = 0 or not, that is, if x (measurement days since cycle start) does or does not contribute information for the prediction of y (boron ppm) over time, given the measurement data available. For this case, the null hypothesis (Ho) is that the slope is actually equal to zero (i.e. there is no trending change in volume over time), the alternative hypothesis (Ha) is that the slope is not zero (i.e. that there is trending change in volume over time). This is a two-tailed test and it uses the following test statistic that follows the student t distribution with DF=N-2:

t= m/SEm and we reject Ho if t < -tcm or t > t,,i at a= 0.05 In the case of the boron regression from U2 SFP, the test statistic and tcH are as follows:c, t = -0.0202 / 0.0288 = -0.7004 tcri(DF=45, a=0.05) = 2.014 Thus, we will reject H0 if t < -2.014 or t > 2.014 And, since the calculated t-value is not less than or greater than tri we do not reject the null hypothesis and conclude that the slope m = 0 in a statistical sense. At the 95%

significance level (a= 0.05), the sample data does not provide sufficient evidence to conclude that the boron concentration (and hence the SFP volume) is changing over time. In fact, this test would show that boron concentration or overall water volume are not changing significantly over time within the boundaries of the error of sampling and analyses of boron measurements performed to generate the regression.

It is instructive to also look at the statistical power (p-value) for this result. Statistical significance is increased as the p-value decreases (i.e. a lower p-value represents a higher confidence in the utility of a test, in this case the regression model). This p-value is compared to our critical level of 0.05 (or 95% confidence).

A t = -0.7004 with DF = 45 gives a p-value = 0.4873.The resulting p-value is much larger than 0.05 and represents a much lower confidence (-51%) in the test than is statistically acceptable. In order to conclude that there is an actual slope (m < 0, or m > 0) we must tolerate too high a risk. Therefore, we must again conclude that the slope m = 0 in a statistical sense.

3 of 5

CONCLUSIONS Alt:hough the regression method results in a calculated loss of volume from the U2 SFP of *-2.6 gpd, the true slope will fall somewhere within the 95% confidence level range for this; prediction and the range of these results is more appropriate to report. In this manner, the regression model predicts that the true loss of volume from the U2 SFP is actually in the range of -9.9 gpd to + 4.8 gpd. As such, the true loss in volume could very well be zero (0) based on this method.

In a test of the utility of the model, it is found that at the 95% significance level (a= 0.05),

the sample data does not provide sufficient evidence to conclude that the boron concentration (and hence the SFP volume) is changing over time. In fact, this test would again show that boron concentration or overall water volume are not changing significantly over time within the boundaries of the error of sampling and analyses of boron measurements performed to generate the regression. The assignment of a hard and fast number for loss of volume from the U2 SFP based on this method and the available data would therefore not be appropriate.

The utility-of the model would certainly be increased if the spread of data were less or if.,;t the actual trend in changing boron concentration were changing much more significantly -

over time.

Wile the strictly rigorous statistical methodologies utilized in this review may differ from the methodology used in the earlier approach, ultimately the results of this review basically agree with the following conclusions made in the "Unit 2 Spent Fuel Pool Total Volume Loss Estimation" memo dated October 27, 2005 (See Appendix A):

  • "chance alone may easily account for the difference in the mean boron concentration"
  • " The normal analytical variances are significant relative to the differences in reported concentrations over the current cycle"; and
  • "The Student's t-test [between early and later means] shows that normal variation (chance) can account for differences in reported concentrations. Therefore, the detection of the true nature of thc present pool loss rate is not as precise as we'd like to have. From the boron data alone, the loss may actually be zero. More data acquisition and analysis is recommended" 4 of 5

awrnit 2 Spent Fuel Pool Total Volume Loss Estimation' October 27, 2005.

Inclian Point Unit 2 Spent FuelPool Investigation Team Review Performed by:

Lawrence T. Dauer, PhD, CHP Date:

_I n!% ., - .. - . -n,* . -. . ,, t s 5 of 5

GRAPH I Unit 2 SFP Boron, Cycle 17 2480 2460 +

2440 E

0.

0.

no 0

0 0

2420 +

In e a)

C, ."I-----

2400 2380 -

2360 0 50 100 150 200 250 300 350 Days Since Cycle 17 Start

APPENDIX A "Unit 2 Spent Fuel Pool Total Volume Loss Estimation" October 27, 2005 Indian Point Unit 2 Spent Fuel Pool Investigation Team A-1

APPENDIX A Oct. 27, 2005 Indian Point Unit 2 Spent Fuel Pool Investigation Team Unit 2 Spent Fuel Pool Total Volume Loss Estimation

Introduction:

The purpose of this evaluation is to provide a reliable computational estimate of the.

current Unit 2 cycle's leakage of the spent fuel pool (SFP). The computation is performed by fitting the current cycle boron concentration function (of time) and calculating the slope. A negative slope indicates a boron loss, a positive slope indicates a boron increase, and a slope of zero indicates no boron change as a function of time. Then the slope (change rate of boron concentration) is divided by the average pool boron concentration to yield the daily boron fractional change rate. This is then multiplied by the SFP volume to cestimate the daily pool volume change rate. Chemistry department analyzes Unit 2 SFP for boron concentration' at about a weekly frequency. There are forty-seven data points of interest at the time the analysis was performed.

Boron is a reasonably good solute to track because it has no reported carryover loss dufing evaporation of water at normal fuel pool temperatures. Its loss can be confidently associated with a volume loss of the SFP inventory. Operations department maintains pool level within a fairly narrow band by periodically adding a makeup that is free of bofic acid. The makeup compensates for evaporative losses.

An additional statistical test (Student's t-test) is also formed to see if late-cycle boron concentrations are different from early-cycle mean boron concentration.

Background:

Possible boron losses and additions of interest from the SFP are:

  • Pool leakage through the liner or a liner penetration, weld or appurtenance
  • Losses through external loop cooling components - such as pump seals
  • Losses through external loop demineralizer components
  • Draining of isolated cooling loop components to perform inspections, tests or maintenance
  • Losses through draining supplemental cooling system components
  • Admixing volumes of similar borated systems - such as the SFP with cooling or demineralizer loops that previously circulated RWST contents

. Planned borated makeup.

There is little hard data on most of these losses and additions. Some notes pertaining to the occurrence of losses or additions do not have accompanying detailed information. For example, System Engineering estimates that mechanical seal losses of the cooling pumps A-2

APPENDIX A are on the order of a drop or two a minute. Operator logs and condition reports were examined for any evolutions or conditions that document intentional system draining, additions containing borated water, or significant degraded component conditions; none were found or could be quantified. There's no data for the amount of pool water that may have been drained from the outage supplemental SFP cooling system when that system was removed. No planned borated makeup activities are noted in this current cycle.

Maintenance replaced a cooling pump mechanical seal package in the August/September timeframe.

Th- Unit 3 water processing operator and a former Unit 2 radwaste supervisor were interviewed to determine what other confounding activities could affect the amount of pool water boric acid concentration. There are some differences in how several of the Unit 2 and Unit 3 borated systems are handled, especially with respect to outage recovery and practices. Neither individual could identify a specific activity that could affect Unit 2 SFP boric acid inventory through draindowns, transfers, or by leakage through pool gates when levels were unmatched.

As a result, no justified adjustment -data:is included that may color the basic computation of SFP boron loss and the leakage estimation that follows it.

Method/Analysis - Basic Loss Computation:

A simple linear regression on the current cycle SFP boron concentration data was composed in EXCEL. There were forty-seven included coordinates (boron concentration, date). The tabular data, statistical computations, and a plotted view for the concentration function are provided in Attachment 1.

Th- mean concentration (47 data points) is 2404 ppm, the y-intercept is 2407 ppm, and the slope is -0.0202 ppm/day. The Standard Deviation (S.D.) of the experimental population of the pool concentration is 18.7 ppin (under assumption of static true concentration). Standard error of the experimental pool concentration mean is 2.7 ppm.

Th. fractional daily boron loss is -.0202 ppm/day / 2404 ppm = -8.4E-6 per day. From layout drawings of the FSB, the average N to S dimension is taken to be 30.7.feet, the E to W dimension is 34 feet and the depth is taken to be 39 feet. The total volume is:

(30.7 x 34 x 39) cubic feet x 7.48 gal/ft3 = 304,000 gallons (+/- 10%). [No fuel and rack volume displacement credited.]

The leakage rate is:

304,000 x -8.4E-6/day = -2.6 gpd.

A-3

APPENDIX A Method/Analysis - Additional Check Using t-statistic Means Testing:

From the point-slope formula, it is evident that the total loss of boron in the pool from the start of the current operating period is about 7 ppm or about 0.3%. This is considerably les3 than the S.D. of the total sampled concentrations in the current cycle. To see if late-cycle concentrations can be thought of as normal variants of the early-cycle concentrations, a Student's t-test was composed.

Th. test uses the forty-seven data points as follows:

  • First sampling is the twenty earliest boron concentrations
  • Second sampling is the last ten boron concentrations
  • The means and their respective standard errors are computed
  • The combined standard error (SE) is computed
  • The degrees of freedom is computed using the Welch adjustment (adjustment for potentially different sample variances)
  • From standard tables, the t-value at 95% CL is selected on the basis of degrees of freedom computed
  • The means difference is compared to the product of the t-value and the SE; if the difference is less than the product so formed, the means are confidently not different from each other and the boron data alone is not suggestive of a boron loss.

Th- t-value and SE product is 10.3 ppm. The difference in actual means is 7.1 ppm; therefore, the means are not significantly different and the boron is not statistically different from the early to latter cycle stages. In other words, chance alone may easily account for the difference in the mean boron concentration.

Conclusion:

Th. boron regression analysis is a proper method for computing boron loss of the SFP. It is selectively affected by loss and addition activities that should be known to the extent practicable. Computation of losses of boron corresponding to more than 10 gallons per day would probably be more reliable than those of a gallon or two per day. The normal analytical variances are significant relative to the differences in reported concentrations over the current cycle.

Thc Student's t-test shows that normal variation (chance) can account for differences in reported concentrations. Therefore, the detection of the true nature of the present pool loss rate is not as precise as we'd like to have. From the boron data alone, the loss may actually be zero. More data acquisition and analysis is recommended.

A-4

APPENDIX A Attachment 1 Unit 2 Cycle 17 Data and Computation

-0.02020 slpe y-Interopt S.D. (ppm)

-0.0202D 2407A2131 18.7 x Y S.E. of Mean

  1. 47 47 (AYW (H) 2.7 1 0 2424 0 0 2 2 2410. 4820 4 3 6 2395 14370 36 Unit 2 SFP BOrOn,ACycle 17 4 13 2392 31096 169 5 20 2456 49120 400 6 27 2396 64892 729 7 42 2413 101346 1764 2450 8 48 2390 114720 2304 9 55 2370 130350 3025 2350 .

10 62 2452 152024 3844

'2300 _3 11 69 2394 185186 4761 12 78 2398 182248 5776 13 83 2392 198536 6889 14 90 2377 213930 8100 2150 15 97 2408 233382 9409 5_

2100 16 104 2407 250328 10816 268176 12321 2050 17 111 2416 18 118 2398 282984 13924 2J000 . t_ 1,,

19 125 2394 299250 15625 20 132 2400 318600 17424 Dat pacycl SrAm 21 139 2406 334434 19321 22 146 2394 349524 21316 23 153 2409 358577 23409 24 160 2432 389120 25600 Vossl 25 167 2393 399831 27889 -2.6 gpd 26 174 2404 418298 30276 27 181 2417 437477 32761 fractional loss per day 28 188 2399 451012 35344 -8.4E-06 29 195 2399 467805 38025 30 202 2411 487022 40804 31 209 2464 514976 43681 32 216 2387 515592 46656 33 223 2401 535423 49729 34 230 2420 556600 52900 dimensions 30.67 N to S 35 237 2390 566430 56169 (fee) 34 E to W 36 244 2411 588284 59536 39 deep 37 251 2412 605412 63001 38 258 2395 617910 66564 esimate frm 39 265 2389 633085 70225 SFP building 40 272 2403 653616 73984 general plan 41 279 2392 667368 77841 42 286 2399 6886114 81796 43 293 2379 697047 85849 44 296 2387 706552 87616 45 300 2403 720900 93000 46 307 2398 736186 94249 47 314 2424 761136 93596 48 0 0 49 0 0 so 0 0 51 0 0 52 0 0 53 0 0 64 0 0 55 0 0 56 0 0 A-5

APPENDIX A 57 0 0 58 0 0 59 0 0 60 0 0 61 0 0 62 0 0 63 0 0 64 0 0 65 0 0 66 0 0 67 0 0 68 0 0 69 0 0 70 0 0 71 0 0 72 0 0 73 0 0 74 0 0 75 0 0 76 0 0 77 0 0 78 0 0 79 0 0 60 0 0 81 0 0 82 0 0 83 0 0 84 0 0 85 0 0 86 0 0 87 0 0 88 0 0 89 0 0 90 0 0 91 0 0 92 0 0 93 0 0 94 0 0 95 0 0 96 0 0 97 0 0 98 0 0 99 0 0 100. 0 0 101 0 0 102 0 0 103 0 0 104 0 0 105 0 0 106 0 0 107 0 0 108 0 0 109 0 0 110 0 0 111 0 0 112 0 0 113 0 0 114 0 0 115 0 0 116 0 0 A-6

q--

APPENDIX A 117 0 0 118 0 0 119 0 0 120 0 0 121 0 0 122 0 0 123 0 0 124 0 0 125 0 0 126 0 0 127 0 0 128 0 0 129 0 0 1[30 0 0 131 0 0 132 0 0 133 0 0 134 0 0 135 0 0 136 0 0 137 0 0 138 0 0 139 0 0 ,.. ...

140 0 0 141 0 0 142 0 0 143 0 0 144 0 0 145 0 0 146 0 0 147 0 0 148 0 0 149 0 0 150 0 0 151 0 0 152 0 0 153 0 0 154 0 0 155 0 0 156 0 0 157 0 0 158 0 0 159 0 0 160 0 0 161 0 0 162 0 0 163 0 0 164 0 0 165 0 0 186 0 0 167 0 0 168 0 0 189 0 a 170 0 0 171 0 a 172 0 0 173 0 0 174 0 0 175 0 0 176 0 0 A-7

APPENDIX A 177 o 0 178 0 0 179 0 0 180 o 0 181 0 0 182 0 0 183 o 0 184 o 0 185 o 0 186 0 0 187 0 0 188 o 0 189 0 0 190 0 0 191 0 0 192 0 0 193 0 0 194 O 0 195 o 0 196 0 0 197 0 0 198 0 0 199 o O 200 o O 201 0 0 sumx4 OuL y1 7465 112998 17938867 1610457 x-bar y-bar sums:

159 2404 A-8

j - -

APPENDIX A Attachment 2 Early Cycle 17 Data and Statistics

-0.106B4 slope y4niercept S.D. (ppm)

-0.10684 2410.83751 21.1 x y S.E. of Mean

  1. 20 20 (xa (x? 4.7 1 0 2424 0 0 2 2 2410 4820 4 3 6 2395 14370 36 4 13 2392 31096 169 5 20 2456 49120 400 6 27 2396 64692 729 7 42 2413 101346 1764 8 48 2390 114720 2304 9 55 2370 130350 3025 10 62 2452 152024 3844 11 69 2394 165186 4761 12 76 2398 182248 5776 13 83 2392 198536 6889 14 90 2377 213930 8100 15 97 2406 233382 9409 16 104 2407 250328 10816 17 111 2416 268176 12321 18 118 2398 282964 13924 19 125 2394 299250 15625 20 132 2400 316800 17424

-13.5 gpd fractional loss perday

-4.4E-05 dimensions 30.67 N to S (feet) 34 E toW 39 deep estimate from SFP building general plan 0.008790004 C.V A-9

APPENDIX A 171 0 0 172 0 0 173 0 0 174 0 0 175 0 0 176 0 0 177 0 0 178 0 0 179 0 0 180 0 0 181 0 0 182 0 0 183 0 0 184 0 0 185 0 0 188 0 0 187 0 0 188 0 0 189 0 0 190 0 0 191 0 0 192 0 a 193 0 0 194 0 0 195 0. 0 196 0 0 197 0 0 198 0 0 199 0 0 200 0 0 201 0 0 11 112 B SAnT~Y 12B0 48080 3073338 117320 x-bae y-be guns:

84 2404 A-10

APPENDIX A Attachment 3 Student's t-test on Cycle Boron Concentrations

.. var S.D. (ppm) s2 var S.D. (pprm 446.5 21.1 147.0 12.1 x Y 8E of Mou X Y S.E. of Mean

  1. 20 20 0.J h? 4.7 # 10 10 3.8 I 0 2424 0 0 1 258 2396 617910 6ta4 2 2 2410 4820 4 2 266 2389 633065 70225 3 6 2395 14370 36 3 272 2403 653610 739S4 4 13 2392 31096 109 4 279 2392 e67386 77841 6 20 2456 49120 400 5 2S6 2399 S68114 81796 6 27 2396 64092 729 6 293 2379 697047 68549 7 42 2413 101346 1764 7 296 23S7 70E552 87515 8 48 2390 114720 2304 8 300 2403 720900 90000 9 5S 2370 130350 =025 9 307 2398 73618S 94249 10 52 2452 152D24 3644 10 314 2424 761136 9850o 11 69 2394 185186 4761 tt a 0 12 76 2308 182248 5776 12 a 0 13 83 2392 198536 6889 13 0 0 14 90 2377 213930 8100 14 0 0 t5 97 2400 233382 8409 15 0 0 16 104 2407 253280 10816 Is 0 0 17 III 2418 288178 .712321 X 17 0 0 1S 118 2398 2S2964 13924 18 0 a 19 125 2394 299250 15626 19 0 0 20 132 2400 318800 17424 20 0 a 21 0 0 21 0 0 22 0 0 22 0 0 23 0 0 23 0 0 24 0 0 24 0 0 28 0 0 25 0 0 26 0 0 28 0 0 27 0 0 27 0 0 28 0 0 28 0 0 29 0 0 29 0 0 30 0 0 30 0 0 31 0 0 31 0 0 32 0 0 32 0 0 33 0 0 33 0 0 34 0 0 34 0 0 3S 0 0 35 0 0 3S 0 0 36 0 0 37 0 0 37 0 0 38 0 0 38 0 0 39 0 0 39 0 a 40 a 0 40 O 0 41 0 0 41 0 0 42 0 0 42 0 0 43 0 0 43 0 0 44 0 0 44 0 0 45 0 0 45 0 0 46 0 0 46 0 0 47 0 0 47 0 0 48 0 0 48 0 0 49 0 0 49 0 a 50 0 0 50 0 0 51 O 0 61 0 0 62 0 0 52 0 0 53 0 0 53 0 0 54 0 0 54 0 0 5S 0 0 55 0 0 56 0 0 56 0 0 A-lI

APPENDIX A 177 o 0 177 0 C 178 o 0 I7T o 0 179 o 0 179 o o 180 0 , 0 180 0 0 181 0 0 181 o 0 182 o 0 182 o 0 183 0 0 183 o 0 184 o 0 184 o 0 185 0 0 1 0 0 186 0 0 186 0 0 187 o 0 187 0 0 188 0 0 188 o 0 199 o 0 189 0 0 190 0 0 190 o *0 191 o 0 191 0 0 192 0 0 192 0 0 193 o 0 193 0 0 194 o a 194 0 0 195 o 0 195 0 0 196 o 0 196 0 0 197 o 0 197 0 0 198 o 0 198 0 0 199 0 O 199 0 0 200 .0'. 0; .. i' 200 0 0 201 0 i"0 201 0 0 stxn X wmm sm xi sum YJ 1280 48080 3073338 117320 23570 2399 8282004 780156 x4war y-bar SUMs x-ber y-or SM 84 2404 287 2397 Sh deris 1-iet SE 6.08 DF- 27.3 t-DF) 1.70 Waih CoSMf udWed dill 10.3 actual dlf 7.1 A-12

Retrieved from "https://kanterella.com/w/index.php?title=ML061000468&oldid=2665640"