ML20209D423

From kanterella
Jump to navigation Jump to search
Actuation Train Relay Response Time Test Summary & Results
ML20209D423
Person / Time
Site: South Texas  STP Nuclear Operating Company icon.png
Issue date: 04/21/1987
From: Friedlander M
HOUSTON LIGHTING & POWER CO.
To:
Shared Package
ML20209D413 List:
References
NUDOCS 8704290269
Download: ML20209D423 (7)
v • d • e


Text

g:f Houston Light'ing and Power Company South Texas Project -

Electric Generating Station Actuation Train Relay Response Time Test Summary and Results Prepared by M.A. Friedlander April 21, 1987

)

s

/

Abstract 2

The response time of selected relays that are typically found in the Actuation Trains of-the South Texas Solid State Protection System (SSPS) was measured. The relays were tested before and after they had been subjected to conditions beyond their qualification.

It was found that the response time of the relays does not significantly change following an accelerated aging proces's.

Furthermore, the contribution of the relays to the total ESF Loop Response' Time in one of the more limiting applications (Feedwater Isolation) represents approximately 0.5% of the total loop response time limit, both before and after the aging process. A limited amount of information was obtained from industry sources. Data representing approximately 350 relay-months of operation over a 2.5 year period was collected and analyzed.

It was concluded that the response time of the actuation train did not change f

beyond statistical variations that are' expected during this type of testing.

Furthermore, the total loop response time did not change appreciably over the testing period, rad under all circumstances, the acceptance criteria was met.

D K

D A

i 3

t t

N Discussion y

During the review of the STP Technical Spe:ifications, the NRC requested HL&P to substantiate a proposed 34 month surveillance interval for Response Time Testing of the ESF Functions'in the STP S Sl'S. HL&P polled the industry and found that very little historical data exists that demonstrates that the response time of the relays in the actuation trains does not change over time.

It was found that since this type of data is not required, it is not typically obtained.

Some of the more recently licensed plants do however, collect this data, and were very cooperative in disseminating the information. In an effort to help evaluate the proposed surveillance interval, data that was available from other plants was collected, and HL&P decided to perform testing to determine whether or not the relays will degrade over time.

The testing that was performed was carried out as follows:

a The first step was to identify the possible failure modes for the relays. The possible causes were determined to be:

o Evaporation of the lubricant on mechanical surfaces o Degradation of the spring action o Degradation of the contact surfaces o Corrosion product buildup Evaporation of lubricant was assumed to be affected by ambient temperature and natural evaporation.

Degradation of the spring action and contact surfaces was postulated to be caused by excessive use.

Corrosion product buildup was assumed to be caused by the ambient environment, specifically, the temperature and relative humidity.

The next step was to establish a testing program ;o svaluate the response of the relays under any or all of the thil re nodes.

The relays are designed to remain functionri I.Jer.he following conditicns:

Normal - Temperature Between 60 and 104 Degrees F and Relative Humidity Between 20% and 70%

Abnormal - Up to 12 Hours With Temperature Between 40 and 120 Degroes F and Relative Humidity Between 0%

and 35%.

It was decided to subject the relays to temperatures and humidity levels outside these limits, and to then measure a response time.

In an effort to induce spring failure and contact failure, the relays were cycled 5000 times (Note: It was determined that a typical relay is cycled approximately 20 times per year as the result of testing and inadvertent actuations).

Response time testing data was collected from a recently licensed plant; this test data was collected over a time period from the Spring of 1984 to Fall of 1986. From another plant, data was collected from the most recent ESF Response Time Test.

The data consists of combinations of total loop and SSPS response times for six ESF functions.

Test Method Baseline data was obtained to determine the initial response time of the system.

One group of relays was subjected to environmental conditions of 127 Degrees F and 90% relative humidity for 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br />.

After this, they were allowed to sit in a normal environment, undisturbed, for 7 days.

The relays were then response time tested over the range of 0-300, 2500-2800, and 5000-5300 cycles. Another set of relays was exposed to an environment of 147 Degrees F and 90%

relative humidity for over 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br />. While the relays were still hot, they were response time tested over the range of 0-300, 2500-2800, and 5000-5300 cycles.

The actual inplant data was analyzed by trending the SSPS response time over the test period.

It was found that the initial test did not differentiate between SSPS response time and total loop response time.

As such, it was necessary to compare the total loop response times over the test period as well as the SSPS reponse time.

Conclusions Table 1 shows the results of the testing performed by HL&P.

Table 2 shows the results of data reduction performed on test results provided to HL&P from other utilities.

By comparing the data collected by HL&P with the two other plants, it was found that the response times for the SSPS are very close.

Af ter being subjected to environmental conditions beyond the limits specified for these components, the response time was not significantly degraded.

It should be noted that prior to the testing, the relays had been in storage at the STP site for over 48 months and an indeterminate time at the manufacturer's warehouse.

Over a 2.5 year period, representing over 350 relay-months of operation, actual inplant data does not indicate degradation of SSPS response time.

l t

Considering the possible failure modes of this type of relay and in light of the extensive testing requirements currently in the Technical Specifications (i.e. Monthly or Quarterly Slave, Master, and Actuation Logic Tests), it is considered unlikely that a failure of the SSPS will go undetected. This type of relay has been used extensively in the nuclear industry with a high degree of reliability. User participation in nationwide programs such as NPRDS and the INPO SOER Program will continue to ensure that any failures that are attributable to these relays are dealt with in the appropriate manner.

1 Table 1 Summary of Relay Response Times For HL&P Testing Initial Response Time (msec)

Group 1 35.6 Group 2

<28.3 Group 1 Relay Response Time After Heating (msec)

(16 Hours at 147 Degrees F, 90% Relative Humidity) 30.05 (0-300 Cycles) 30.07 (2500-2800 Cycles) 30.05 (5000-5300 Cycles)

Group 2 Relay Response Time After Heating (msec)

(16 Hours at 127 Degrees F, 90% Relative Humidity, Followed By 7 Days Undisturbed) 35.3 (0-300 Cycles) 34.2 (2500-2800 Cycles) 34.7 (5000-5300 Cycles)

~

Table 2 i

Comparison Of' Response Time Data 1

4 SSPS Response Time STP Data: Average Response Time - 32.3 mSee Plant 1: Average Response Time (6 Functions)- 26.0 msec Spring 85 - 24.7 msec Winter 86 - 22.7 msec t-Fall 86

- 26.1 msec (1)

Plant 2: Average Response Time - 32.5 msec i

i i

Total ESF Loop Response Time Plant-1: Average ESF Loop Response Time (Low Pressurizer Pressure SI)

(Acceptance Criteria - Loop Response Time <2 Sec)

Spring 84 - 71.4 msec f

Spring 85 - 67.0 mSee Winter 86 - 69.2 msec Fall 86

- 113.5 mSee l-i 1

7 7

(1) Steam Line Pressure Low Data Excluded From This Average Due To

[

Suspect Data 2

I

.~.

ATTACHMENT 2

~

ST.HL AF..All a '

.l5[ffnr 106nkdd)-,

-- - ~ mvr wr i INSTRUMENTATION

/

\\

SURVEILLANCE REOUIREMENTS 4.3.2.1 Each ESFAS instrumentation channel and interlock and the automatic actuation logic and relays shall be demonstrated OPERABLE by performance of the ESFAS Instrumentation Surveillance Requirements specified in Table 4.3-2.

4.3.2.2 The ENGINEERED SAFETY FEATURES RESPONSE TIME of each ESFAS function shall be demonstrated to be within the limit at least once per 18 months,_

Each test shall include at least one train so that9all trains are tested at G ast once per 36 months and one channel per function so that all channels l

are tested at least once per N times 18 months where N is the total number J of redundant channels in a specific ESFAS function as shown in the "T jtal o

(No. of Channels" column of Table 3.3-3. f gheuh\\o1akr^hhYe5kdE\\c^5k*^CEP"'. Mn. a b+d & hm

"^

I ac w y c.6 ne\\ p.t b e h so % & J chmswe\\s are k h d. mt \\ cut o %

Pec 4%u 19 mcnk & N n h tot ) nwnbe.t of celundegb c6nels rn o sp<ch 33m3 G%n ms sgcen in % wnt4 Mo. e4 coIu mn ok D}c

~1.'3-3,

"*"t IA D O

l l

l O

SOUTH TEXAS - UNIT 1 3/4 3-17

_ _. _ _

Retrieved from "https://kanterella.com/w/index.php?title=ML20209D423&oldid=4028554"