ML19319B467
| ML19319B467 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 08/03/1977 |
| From: | Roe L TOLEDO EDISON CO. |
| To: | James Keppler NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| References | |
| NUDOCS 8001220875 | |
| Download: ML19319B467 (4) | |
Text
-
-. ; ;., c., '
- c a
e telecopied to:
R. Knop '
K. Seyfrit TOLEDO 8/3/77 D. Vassallo Ei[alEiCri'J Docket No. 50-346 LOWELL E. ROE
%ce Prevdent Au8ust 3, 1977 me> ass sa4a Serial No. 339 Mr. James G. Keppler Regional Director-Region III Directorate of Regulatory Operations Nuclear Regulatory Commission 799 Roosevelt Road Glen Ellyn, Illinois 60137
Dear Mr. Keppler:
This letter is in response to an oral request by Mr. Karl V. Seyfrit of the NRC Office of Inspection and Enforcement, Division of Reactor Inspec-tion Programs, regarding the Davis-Besse Nucle =- Power Station Unit No. I startup and test program.
With regard to Pre-Operational Test IP 401.11, Instrument Ground Grid i
System Test (reference item (2) of NRC letter of April 8, 1977 to me from Mr. John F. Stolz), we have been unable to complete the test as originally anticipated. This test was developed as a part of the over-all testing program for Davis-Besse Unit 1, and was not developed as a result.of any NRC requirements nor was it included in the testing require-ments set forth in Chapter 14 of the Davis-Besse Unit 1 FSAR. Your inspectors, however, had re.uested from our facility operating personnel that the test be subject to their review, and as a result it was included in our test program relief request letters to NRR dated February 22, 1977 and March 2, 1977.
This letter addresses the bases for our design intentions regarding instrument and station grounding, identifies why the anticipated testing cannot be completed as planned, and discusses the acceptability of the installed grounding system at Davis-Besse.
The original design objective for Davis-Besse Unit I was to have an isolated instrument ground system that would be connected to the station grcund system at ont point only. We soon found that equipment to be supplied by Babcock & Wilcox Company (B&W) and Bailey Meter Company was not designed to have a separate instrument and station ground, but rather only a single station ground.
197I AMG 8 hk b THE TOLEGO EDISON COMPANY EDISON PLAZA 300 MAO! SON AVENUE TOLEDO. OHIO 43652 soo3220f 75 g #
.~
i As expected, testing of the instrument ground system revealed that the j
instrument ground was connected to the station ground at more than just i
one point. A cabinet by cabinet test of ground isolation confirmed that the following systems did in f,ct have an isolated instrument ground as i
designed:
Safety Features Actuation System Channels 1, 2, 3 and 4;
}
Steam and Feedwater Rupture Control System Channels 1 and 2; Turbine l
Supervisory Instrumentation; Meteorological Station Panel; Start Up Test Panel; and Switchyard Control Console. The following systems have the station ground and instrument grounds connected: Nuclear Instrumentation-Reactor Protection System, Non-Nuclear Instrumentation, Station Computer, Integrated Control System, Control Rod Drive System, and the non-essential l
Miscellaneous Electric Control System, which are all B&W or Bailey supplied equipment.
At this point, let us identify some of the features of various grounding approaches.
j I
1)
Single Station Ground (SG) System Such a system comprises the structural steel in the station buildings, the ground grid throughout the plant, and all the grounds connected j
to the facility power equipment, switchgear, enclosures, racks and j
panels. This system firmly connects the site to the earth.
A perturbation in the SG System (e.g., a faulted motor) can cause l
currents to flow throughout the SG grid.
Because this is a highly interconnected grid, predicting exact current flow is impossible.
Current flow can cause different potentials to appear at points
}
along the SG grid.
If an instrument circuit ground is connected to j
the SG at two dif ferent points, they may be at two lifferent i
potentials. This difference in potential could cause current to flow in the ground circuit of the instrument; this unwanted current j
flow can also be considered as n 'ise, but with proper design consider-i ations, such noise or unwanted c rrent flow will not affect proper system operation.
2)
Two Ground System: Station Ground (SC) and Instrument Ground (IG)
In addition to the conventional station ground (SG) ; rid system, l
this system also has an instrument ground which is usually an insulated copper bus to which instrumentation and control syst a i
analog and digital instrument grounds attach. This bus is connected i
j directly to the station ground grid at one point wich an insulated to l
copper cable. A perturba ion in the SG System can cause current flow around the point where the SG and IG are connected. This current flow can cause a change in potential at that point. Because instruments are eventually tied to that point via the instrument l
ground, they will all raise and lower in potential together, thus j
eliminating the chance of any im.rument being grounded at two j
different station ground potentials. Because of the size and I
l complexity of the SG System the actual relative potential cannot be l
calculated. Since there is only one connection to SG, ground loops from the SG to the IG are eliminated. There is nothing critical as to where the tie is to be made between IG and SG as long as it is a l
good connection to the SG grid and it occurs at one point only.
i j
The above descriptior represents an ideal two ground system which is rarely, if ever, achieved in a complex industrial installation.
For example, the integrated effect of instrument system cable insulation reduces the isolatien effectiveness. Also, in complex installations consisting of mult:ple suppliers using ditferent grounding techniques, complete isolation may not be attainable while capitalizing on other system attributes.
3)
Two Ground System: IG and SG With Tie (s) Between the Two A tie between the IG and SG may occur by having an instrument tied at one point on the IG and another part of its circuit tied to the SG.
Because the IG is eventually tied to the SG this is the same as having the instrument tied to two points on the SG as explained in 1) above.
Because of possible different potentials between the SG and IG, a current loop can occur through the cable that ties the IG to the SG.
This current flow might affect only those instruments that are part of the tie.
Ihe potential of the instrument ground bus may be different than at the IG to SG tie, but it does not affect the IG isolation of other instrument systems because in effect the ground point of these instrumen s is now merely shifted to a new coc=on location in the IG System; no ground loops can occur in the other equipment.
A cross tie has the effec; of returning that portion of the instrument ground system to a single station ground system. This tie would usually be of a high impedance (i.e., cable shield) when compared to the station ground grid impedance (heavy gauge copper cable),
I thus limitir.g the amount of current through the tie path.
j As I indicated earlier, the design objective for Davis-Besse Unit I was to provide an isolated instrument ground system connected to the station I
ground grid at one point only, arrangment 2).
However, arrangement 3) is the end product of our design. At the time the B&W rystems were l
designed it was not industry practice to provide separate instrument l
grounds. The need for a separate grounding system is dependent upon the The signal-to-noise ratio required for satisf actory system operation.
4 existing systems utilize centrol signals which would require relatively j
powerful long duration noise signals before interference with proper l
operation would be affected.
It has bean demonstrated that the grounded In addition, systems employed on Davis-Besse 1 are relatively noise free.
i the instrumentation systems are designed to process the control signals and not produce unwanted act ons or spurious noise by employing shielding, filtering and transient suppression devices and separation. The noise rejection designa employed ensure satisfactory s stem operation without an isolated instrument ground. Satisfactory system performance has been demonstrated through more than 10 years of reactor operating experience.
me.
_m,-
~
~
4_
The Safety Features Actta' tion System and the Steam and Feedwater Rupture Control System, as well as the other identified isolated systems, use sophisticated solid state circ.titry, and are designed to provide a separate instrument ground bus isolated from the cabinet steel and the station ground grid. The instrument ground bus is connected to the station ground grid at one point only by means of a 500KCMIL connection.
The performance of the instrumentation system during the Hot Functional Test and subsequent Pre-Operation Tests, including facility operation in Modes 6 through 3, has shown that no adverse effects or malfunction of the instrumentation and control systems or equipment due to the grounding systems occur.
The Test Procedure TP 401.11 cannot be completed as intended since its objective was to measure the isolation resistance between the instrument and the station grounding systems when it was assumed this feature would prevail throughout the plant. With the B&W/ Bailey supplied instrumenta-tion not so designed, the test would not be meaningful, therefore, it will not be completed.
Yours very truly, WFW O-C.
&J Lowell E. Roe Vice President Facilities Development db c/1-4 i
t b
r
_